Introduction

Due to its excellent thermal and electrical characteristics, copper (Cu) is a widely used operational material in electronic devices, energy conversion, and heating and cooling exchangers in industries (Ref 1). However, the oxidation of Cu due to its chemical reactive nature is a serious reliability issue in micro/nano-electronics that rendered its applications limited (Ref 2). This problem is more serious when Cu is in close contact with an environment having chloride ions (Ref 3). Therefore, the development of an efficient protective coating is essential against oxidation and corrosion in order to realize a wide range of applications of Cu.

Recently, researchers working in the field of materials science have shown the potential of graphene as a surface protective coating on metals against corrosion species. This corrosion protection behavior of graphene is related to some of its characteristics such as toughness, hydrophobicity, and impermeability to molecules (Ref 4,5,6), making it a better candidate for an ultrathin protective coating for Cu against marine and saline environment (Ref 7,8,9). Graphene coating on metals has proven to be a natural diffusion barrier to aqueous corrosion reactants (Ref 5).

Some researchers reported that graphene could efficiently protect Cu against electrochemical degradation in aqueous media (e.g., NaCl and Na2SO4) by reducing the corrosion rate significantly compared to bare Cu (Ref 7, 10, 11). In these studies, graphene was deposited on a Cu substrate through chemical vapor deposition (CVD) method. The CVD method is known to produce a large area uniform graphene film on metals (Ref 12, 13). It is worthwhile to mention that the characteristics of graphene can be influenced greatly by CVD parameters, such as gases (argon, hydrogen, and methane), temperature, annealing time, and growth time (Ref 8, 14,15,16,17,18,19,20). Our previous study showed that hydrogen concentration during annealing stage affected both graphene characteristics and surface morphology of underneath Cu (Ref 14).

Though a number of corrosion studies were reported after growing graphene on Cu, to the author’s knowledge, no systematic study was performed when graphene was grown by CVD method under different hydrogen concentrations during annealing stage. In this paper, we performed a systematic corrosion study on Cu after growing different layers of graphene on it. The graphene was grown by altering hydrogen concentration between 0 and 50% of the total gas flow during the annealing stage in the CVD reactor. The chemical structures of graphene/Cu were analyzed by Raman spectroscopy, and the microstructural features of graphene/Cu surfaces were examined by scanning electron microscopy (SEM), transmission electron microscopy (TEM), and atomic force microscopy (AFM). The electrochemical corrosion performance of graphene/Cu was studied in sodium chloride solution (0.5 M NaCl) using potentiodynamic polarization and electrochemical impedance spectroscopic measurements. The corrosion rates were calculated using Tafel analysis. The obtained results could be useful for several applications where the surface of Cu is in close contact with corrosion species.

Experimental Work

Synthesis of Graphene Films

In this study, graphene films were grown on thin Cu foil (Alfa Aesar No. 13382) using different hydrogen (H2) concentrations during the annealing stage of the chemical vapor deposition (CVD). A custom-made atmospheric pressure CVD system (First Nano Co. USA) was used for growing graphene and annealing of Cu specimen (reference sample). A standard CVD cycle includes four main steps in the following order: (1) preheating, (2) annealing, (3) growth, and (4) cooling. Notably, to study the effect of H2 on graphene and Cu characteristics and hence on corrosion properties of Cu, the concentration of H2 was varied during step 2 (annealing step) only, while the ratio of H2 and argon (Ar) was kept constant (37:1463) during steps 1 and 4 for all runs.

At first, Cu foil with dimensions 1 × 4 cm2 was heated (from room temperature to 1000 °C) in Ar and H2 gas mixture environment with a total flow rate of 1500 sccm. In the second step, Cu foil was annealed at 1000 °C for 30 min under H2 concentrations of 0, 2.5, 20, and 50 vol.% of the total gas mixture. Graphene was grown during the growth process by flowing methane gas into the system chamber for 3 min with a flow rate of 5 sccm. Graphene/Cu specimens were then cooled down to room temperature to complete the CVD process under H2 and Ar gas mixture (37:1463). The details of the CVD method are reported elsewhere (Ref 14). Depending on H2 concentration (0, 2.5, 20 and 50%H2), the graphene samples were denoted as G-Ann-0%H2, G-Ann-2.5%H2, G-Ann-20%H2, and G-Ann-50%H2. The reference Cu sample with no graphene (annealed Cu) is labelled as Ann-Cu.

Characterization of Graphene Films

After synthesis, annealed Cu in the CVD chamber without growth (Ann-Cu) and G/Cu specimens were tested for corrosion using potentiodynamic polarization experiment. The morphological characteristics of the as-grown graphene coatings were analyzed by field-emission scanning electron microscopy (FE-SEM) (TSCAN-MIRA 3 LM-USA), and TEM (JEM-2100F, Japan) and AFM (Dimension Icon-Bruker). The structural analysis was carried out by Raman spectroscopy with laser wavelength of 455 nm and optical microscope (DXR™ Raman Microscope, Thermo Scientific, USA). For TEM examination, graphene flakes were transferred onto TEM grids having holey carbon support film after etching the underneath Cu substrate. The SEM and TEM instruments were used at accelerating voltage of 5 and 200 kV, respectively.

Electrochemical Tests of Graphene/Cu Specimens

The specimens were tested for corrosion in a three-electrode cell assembly in the following order: (1) Cu specimens, (2) platinum wire and (3) a saturated calomel electrode (SCE) acting as a reference electrode. The electrochemical corrosion experiments were completed by exposing an area of 0.2 cm2 of the specimen in 0.5 M sodium chloride (NaCl) solution. The corrosion data were collected using Gamry Reference 3000 Potentiostat electrochemical workstation. Potentiodynamic polarization (PD) tests were carried out at ± 250 mV with respect to open-circuit potential with a scan rate of 0.2 mV per second. Corrosion parameters were obtained by analyzing the Tafel regions by applying DC105 corrosion software. Electrochemical impedance spectroscopy (EIS) experiments were performed in the frequency region of 0.10-100,000 Hz with a signal amplitude of 0.10 V. In order to verify the reproducibility of the obtained results, the electrochemical tests were repeated at least three times. The Echem Analyst software was employed to study the EIS results.

Results and Discussion

Morphological Analysis of Graphene/Cu Specimens

To study the corrosion behavior of Cu, different types of graphene films were grown on Cu substrate by altering the H2 concentration during annealing stage in the CVD reactor. Since during the wet corrosion tests the surface of the specimen makes close contact with a solution, it is crucial to evaluate the surface morphology of the G/Cu specimens prior to any tests. The surface morphological features and the thickness variation of the graphene/Cu were characterized using SEM and AFM (Fig. 1). The electron micrographs of the G/Cu specimens show the different types of deposited graphene (Fig. 1a, b, c, and d). The change of contrast throughout the electron image reflects the thickness variation in the film. The dark regions on a bright background are the small graphene flakes, denoted here as graphene domains. Interestingly, these domains were observed on all specimens; however, they were varied in size, density, and thickness. For example, thicker and large-sized graphene domains were observed for G-Ann-20%H2 and G-Ann-50%H2 specimens (Fig. 1c and d). By seeing the dark contrast of the domains, they are relatively thicker than the domains of remaining two specimens, G-Ann-0%H2, and G-Ann-2.5%H2. The deposition of large-sized and irregular-shaped multilayer graphene domains in case of higher H2 is attributed to disassociation of high amount of CH4 gas (used as carbon source) and the presence of high-density nucleation sites (Ref 15).

Fig. 1
figure 1

Morphology of graphene films over Cu substrate. SEM images of G/Cu specimens annealed at different H2 concentrations (a–d). AFM micrographs of G/Cu specimens (e–h) along with their roughness profiles (i–l). The roughness profile was taken from the area as marked with a line in each case (avoiding grain boundaries and particles to extract roughness information only). Note: The value of RMS was reduced from 12.6 to 3.6 nm with increasing H2 concentration

Full size image

Further to graphene morphology, Cu exhibits a high density of steps on its surface when it was annealed under 0%H2 condition as depicted clearly by AFM (Fig. 1e). These Cu steps can cause extrinsic defects in graphene films and affect the continuity of the deposited graphene around these features (Ref 14, 15). Remarkably, the height of the steps was reduced significantly when Cu was annealed under high H2 concentration (Fig. 1h), consistent with earlier work (Ref 16). AFM results revealed that Cu surface became smoother by increasing H2 amount during annealing. The root-mean-square (RMS) value was reduced to 3.6 nm (G-Ann-50%H2) from around 12.6 nm (G-Ann-0%H2). Instead of steps, the surface of Cu annealed under high H2 exhibited bigger particles (Fig. 1g and h) compared to G-Ann-0%H2; more details on this aspect are reported elsewhere (Ref 14, 15). It should be noted that the surface of the as-received Cu foil often shows high surface roughness due to the presence of strain lines (rolling lines) and grain boundaries. These surface features are diminished to some extent when Cu is annealed under different H2 conditions during the CVD process. A step-like structure is formed instead if Cu is annealed without H2 due to oxygen impurities present in the Cu foil as a result of Cu reconstruction (Ref 21). On the other hand, Cu surface becomes smooth and flat when annealed under H2 due to a reduction effect of rolling lines. This kind of surface could be beneficial for the growth of a uniform layer of graphene and hence gives better surface coverage. From SEM and AFM analyses, it is concluded that H2 plays an important role in defining the surface morphology of Cu and subsequently the growth of graphene.

The growth mechanism of graphene under different H2 concentrations is elaborated in detail elsewhere (Ref 14, 22, 23); herein, a summary of this mechanism is provided. For low H2 (0 and 2.5%H2), the surface-active sites, e.g., impurity particles, Cu steps, and rolling lines are oxidized and passivated and hence resulted in fewer nucleation sites. Due to the availability of fewer surface-bound active H2 atoms (-H2*), less adsorption and dissociation of methane takes place. The presence of fewer C atoms for nucleation points results in slow growth rate of graphene. For higher H2 (20 and 50%H2), the surface of Cu is very active leading to an increase in the density of -H2*, causing stronger adsorption and dissociation of methane (presence of more carbon) along with a high density of nucleation sites. Consequently, the growth rate of graphene is high yielding to the formation of multilayer graphene.

Structural Analysis of Graphene Films

The thickness and structural information of graphene film on Cu were analyzed by Raman spectroscopy, as shown in Fig. 2a. The graphene films are displaying G (1570-1600 cm−1), 2D (2690-2740 cm−1) and D (1350-1370 cm−1) bands, the characteristics of graphene with minor defects. The intensity ratios of 2D- and G-bands are estimated higher than 1 (I2D/IG ≈ 1.5) for G-Ann-0%H2 and G-Ann-2.5%H2 specimens suggesting growth of mono/bilayer graphene films (Ref 14, 17). In contrast, multilayer graphene films were identified for G-Ann-20%H2 and G-Ann-50%H2 when a higher amount of H2 was used as judged by I2D/IG ratio (≈ 0.7), being less than 1 for multilayer graphene (Ref 11, 24). Moreover, it was noted that the graphene films grown under high H2 concentration displayed a rather high D-band, suggesting presence of defects in these films. It is known that D-band is originated due to several factors, and formation of high-density multilayer graphene domains in the current case could be one of them (Ref 18, 25, 26). Raman spectroscopy results suggested that annealing Cu under Ar or low-pressure H2 content produced fewer layers graphene as compared to the excessive amount of H2. These results are consistent with earlier works (Ref 18, 19).

Fig. 2
figure 2

Structural analyses of graphene films using Raman spectroscopy and TEM. (a) Raman spectra of G/Cu samples showing the characteristic bands, 2D and G of graphene film, and D-band basically displaying the density of multilayer graphene domains. (b) A typical TEM image of graphene over a holey carbon film (G-Ann-20%H2 specimen was considered for TEM). (c) HR-TEM of the graphene taken from the area as marked with a black arrow in b. Inset is the intensity profile taken from the area highlighted by a black line showing 4-5 layers of graphene

Full size image

Graphene film of G-Ann-20%H2 specimen was further analyzed by TEM (Fig. 2b and c). Figure 2b is a typical TEM image of graphene over holey carbon film. A high-resolution image of a graphene edge is displayed in (Fig. 2c), showing 4-5 layers of graphene with a constant distance of around 0.34 nm, corresponds to the distance between any two graphene layers (Ref 27, 28). Indeed, TEM analysis supports the Raman results presented in Fig. 2(a), confirming that G-Ann-20%H2 specimen has multilayer graphene film.

In summary, SEM, AFM, and Raman results have demonstrated that hydrogen concentration during annealing stage has influenced the characteristics of grown graphene in terms of film thickness, and uniformity, as well as surface morphology of Cu. A high H2 concentration during annealing promotes high density of multilayer graphene domains on a uniform graphene film, while a low amount of H2 exhibits fewer domains. A thicker graphene film can be obtained on a smooth surface if the Cu is annealed under high H2 concentration.

Electrochemical Corrosion Analysis of Graphene/Cu Specimens

Figure 3(a) displays the potentiodynamic polarization (PDP) response of annealed Cu and graphene-coated Cu specimens. In PDP curves, it was observed that the corrosion current density (icorr) of the G-Ann-0%H2 specimen was lower than that of annealed Cu, suggesting that the graphene coating on Cu decreased Cu dissolution. Another observation was the increase in corrosion potential (Ecorr) of graphene-coated specimen. Ecorr of the G-Ann-0%H2 specimen was 53 mV more positive compared to the annealed Cu. The shift in Ecorr toward noble direction means an increase in corrosion resistance due to graphene coating. The obtained corrosion behavior of graphene coating on Cu is consistent with earlier work reported by Sing Raman et al. (Ref 10).

Fig. 3
figure 3

Corrosion study of Cu substrates. Potentiodynamic polarization (PDP) response of annealed Cu and graphene-coated Cu specimens exposed in 0.5 M NaCl solution at room temperature

Full size image

For G-Ann-2.5%H2 specimen (blue curve), Ecorr increased compared to G-Ann-0%H2 specimen, and cathodic reaction rate also decreased. This means that the addition of graphene contributed to the decrease in both the cathodic and anodic reaction rates simultaneously. Similar behavior was observed for other specimens, such as G-Ann-20%H2 and G-Ann-50%H2; however, the effect was predominant for this Cu (see red and green curves in Fig. 3a). Some previous work also reported this kind of behavior of graphene-coated specimens. For instance, Kirkland et al. (Ref 7) deposited graphene film on Ni and Cu and showed that the films behaved differently on these two substrates. For Ni, graphene mainly inhibited the anodic dissolution of Ni, while the effect was more significant on the cathodic reaction rates in Cu. However, in both cases, the corrosion rate decreased because of the graphene film.

In the present work, graphene contributed to the inhibition of both the anodic and cathodic reaction rates compared to annealed Cu. Remarkably, the effect was more prominent for G-Ann-20%H2 and G-Ann-50%H2 compared to G-Ann-2.5%H2 and G-Ann-0%H2 specimens, suggesting that graphene film deposited under these conditions was more stable and protective. The coating was protecting the underlying Cu surface by minimizing the movement of ionic species between the solvent and the substrate. The decrease in icorr and increase in Ecorr indicate that graphene is mainly acting as an anodic barrier for these specimens, which ultimately decreases the corrosion rate. The comparison of corrosion data, Ecorr, icorr, ipass (passive current density), and Epit (pitting potential), measured at − 200 mV for all specimens is given in Table 1. Data of this table show that the lowest icorr was observed for G-Ann-20%H2, indicating that this specimen has the lowest corrosion rate as confirmed by Tafel analysis.

Table 1 Corrosion data (measured at − 200 mV) of annealed Cu and G/Cu. The corrosion rate was estimated in millimeter per year (mpy)
Full size table

In order to get more insight on the corrosion protection and electro-kinetic behavior of graphene-coated Cu specimens, nondestructive and conventional EIS measurements were performed in 0.5 M NaCl solution. The obtained EIS data are presented in Fig. 4a in Bode format (impedance module and phase angle vs. log (freq.)). The data are investigated using EIS fitting analysis using the equivalent circuit and displayed in Fig. 4(b), and the relevant EIS values are listed in Table 2. Rs and Rct signify the electrolyte and charge transfer resistance, respectively. Constant phase element (CPEdl) denotes constant phase element of the double-layer capacitance of the barrier layer. In order to achieve precise fitting results, capacitive element (C) was substituted by CPE and the impedance of a CPE is described as ZCPE = [Y0 (j ω)n]−1, where Y0 and n mention frequency-independent values and ω represents the angular frequency (Ref 29, 30). The value of n fluctuating from 0 to 1 has been intensely governed by the surface microstructure, where a value of 1 designates that the surface is perfect and that the impedance of the CPE is deliberated as a pure capacitor (Ref 31).

Fig. 4
figure 4

Bode plots of annealed Cu and graphene-coated Cu specimens exposed in 0.5 M NaCl solution at room temperature

Full size image
Table 2 EIS parameters for the annealed and graphene-coated Cu in 0.5 M NaCl solution
Full size table

From Table 2, the Rct values obtained for G-Ann specimens were comparatively higher than that of annealed Cu specimens, corresponding to the physical barrier behavior of graphene coatings (Ref 32). The G-Ann-20%H2 specimen exhibited higher Rct value (7288 × Ω cm2) than their respective G-Ann-Cu, demonstrating the improved surface protection effectiveness of graphene layer with 20% concentration of H2, probably owing to a strong adherent, different film morphology with less porousity in nature (Ref 33). It is evidently observed that the values of CPEdl were significantly reduced due to the presence of the graphene layer which inhibited the pathway of aggressive elements from the electrolytes toward Cu surface (Ref 34). In particular, compared with the CPEdl value of annealed and graphene-coated Cu specimens, G-Ann-20% specimen exhibits the lowest CPEdl value (0.56 µFcm2), indicating that the graphene layer processed under this condition possessed a stable film/Cu interface with no sign of corrosion damages. The exponent of CPEdl, ndl, of investigated G-Ann-Cu substrates is almost nearing to 1, indicating that graphene film processed at annealed Cu specimen was more compact with a homogeneous surface than that of annealed Cu (ndl-0.91), and the enhanced barrier performance of graphene film efficiently hindered the diffusion of aggressive species from the electrolyte (Ref 35). Based on the obtained results from electrochemical corrosion studies, the surface protective behavior against corrosion provided by graphene coatings is categorized as G-Ann-20%H2 > G-Ann-50%H2 > G-Ann-2.5%H2 > G-Ann-0%H2 > G-Ann.

In summary, the corrosion results revealed that a change in H2 concentration did affect the quality of deposited graphene which in turn affected its corrosion performance. The best results in terms of anodic/cathodic inhibition were obtained when graphene film was grown after annealing at 20 and 50%H2 concentration, i.e., G-Ann-20%H2 and G-Ann-50%H2 specimens. This corrosion behavior could be correlated with the graphene characteristics and the surface morphology of underlying Cu. It is shown in Fig. 2 that graphene film varied in thickness and quality when produced at different H2 concentrations. Depending on the annealing conditions, bilayer and multilayer graphene films were grown. Though the films produced at 0 and 2.5% of H2 had a lower Raman D-band, however, they showed lower corrosion resistance compared to 20%H2 specimen (Table 1). This could be understood as follows: at 0 or 2.5%H2, mono/bilayer graphene film was obtained along with several oxide particles. In addition, Cu surface exhibited a high density of steps and thus high roughness (see Fig. 1e). As a result, graphene grown on such high surface features was unlikely to cover the surface completely and hence the enhancement in corrosion resistance was limited. On the contrary, multilayer graphene grown over Cu under 20 and 50%H2 concentrations was able to decrease the corrosion rate substantially when compared with annealed Cu. In this case, the Cu surface was smooth without any steps as displayed by AFM images (Fig. 1g and h). The graphene grown on such surfaces had better coverage and uniformity, and hence better protection against corrosion as dictated by Tafel analysis (Table 1).

Surface Morphological Analysis of Corroded Specimens

To observe the corrosion effects, optical micrographs and Raman spectra of the annealed Cu and four G/Cu specimens (G-Ann-0%H2, G-Ann-2.5%H2, G-Ann-20%H2, and G-Ann-50%H2) were captured before and after the potentiodynamic test as shown in Fig. 5. After corrosion tests, the Raman spectra of the corroded specimens at different regions were also taken, and the results of all G/Cu specimens are shown in Fig. 5(i), (j), (k), and (l). No major difference in surface morphology of G/Cu specimens was observed before the corrosion test (Fig. 5a, b, c, and d). However, the Cu surface exhibited a totally different morphology and colors after corrosion tests as shown in Fig. 5(e), (f), (g), and (h). Corrosion effects were seen on all four specimens with some differences. The common colors were blue, yellow, and green. The blue color was dominating in case of G-Ann-50%H2 specimen, and bands of blue color were seen along with yellow traces. Raman spectra on each color (of specimens e-h) were taken to find out the presence of graphene (if any) after corrosion test, as shown in Fig. 5(i) and (l). The blue regions in each specimen showed the characteristic features of graphene, D, G, and 2D bands, indicating that graphene was present in this region even after the electrolyte test. On the other hand, the graphene peaks of the Raman spectra disappeared or showed negligible intensities at green and yellow regions (see green and yellow spectra), suggesting that the graphene was detached from these regions during the attack of corrosion species. The additional peaks in the range of 1000-1300 cm−1 were attributed to the fluorescence of Cu (Ref 36).

Fig. 5
figure 5

Surface topography of Cu substrate and graphene film characteristics after corrosion experiments. Optical microscopy (1000 ×) of G/Cu specimens (a–d) before and (e–h) after potentiodynamic test. The topological images of Cu are showing the obvious corrosion effects after corrosion test as depicted by different colors. The corresponding Raman spectra of the specimens (i–l). The spectra were extracted by selecting an area on each color and thereafter represented with that color. The blue area in each specimen is showing the characteristic features, D, G and 2D bands of graphene. Note: The Raman spectra of the specimens (a–d) are shown in Fig. 2(a)

Full size image

The surface morphology of the corroded G/Cu samples was also examined by SEM. Figure 6 shows the SEM micrographs at two magnifications of the graphene-coated Cu specimens after electrochemical tests. Note, the same specimens are shown in Fig. 1(a), (b), (c), and (d) before corrosion experiments. The Cu surface clearly demonstrated the pronounced corrosion when graphene films were grown at low H2 concentration. The G-Ann-0%H2 specimen severely corroded as evident by the cracks and corrosion pits; both small and deep pits were observed suggesting that the graphene film was not continuous due to Cu-stepped structure. The occurrence of deep pit sites (white arrowhead) is attributed to the discontinuous graphene layer thus allowing the electrolyte to pass through it. The G-Ann-2.5%H2 specimen showed a little better protection but its surface was yet completely perturbed. This little improvement in the corrosion performance was attributed to uniform coverage of the graphene due to the smoother surface of Cu produced under 2.5%H2. In contrast, G-Ann-20%H2 and G-Ann-50%H2 specimens showed better protection against corrosion. Only few localized corrosion sites were distributed on the surface (Fig. 6f and h), indicating that the Cu surface was well-covered and protected by a graphene film. These SEM images further support the corrosion results shown in Fig. 3, confirming that Cu with multilayer graphene exhibits a better corrosion performance.

Fig. 6
figure 6

Surface morphology of Cu substrate after corrosion tests. SEM micrographs of the corroded surfaces at two magnifications; (a, c, e, g) large area and (b, d, f, h) close-up view of the specimens after corrosion tests. The surface of the G-Ann-0%H2 specimen was severely damaged compared to other specimens due to attack of corrosion species

Full size image

The better corrosion performance of G-Ann-20%H2 and G-Ann-50%H2 specimens is ascribed by two factors which are closely linked to each other, i.e., smooth surface of the Cu and the growth of multilayer graphene. The smooth surface that is absence of Cu-stepped structure, and larger Cu grains are beneficial for the growth of a uniform layer of graphene. The multilayer graphene reduces the passage of corrosion species considerably, and hence the corrosion rate, as reflected in the electrochemical results as given in Table 1. Graphene deposition improves the corrosion resistance of Cu, such that the corrosion rate of Cu (Ann-Cu) is 5.2035 mpy, while the coated specimen (G-Ann-20%H2) results in 0.0087 mpy, nearly three orders of magnitude slower than the annealed Cu. Comparing the coated specimens with uncoated ones, the surface morphology of the bare Cu (without graphene) was also evaluated by SEM before and after corrosion experiments (Fig. 7). The surface of Cu severely corroded after being subjected to corrosion species. The untouched Cu showed a smooth surface with some grains.

Fig. 7
figure 7

SEM micrographs of the bare Cu (a) before and (b) after potentiodynamic tests. The surface of the Cu was severely damaged after corrosion

Full size image

Corrosion Mechanism

The possible corrosion mechanism is shown in Fig. 8. SEM and AFM analyses reveal that Cu exhibits a rough surface when annealed under zero and/or low H2 content during the annealing process, while showing a smooth surface at higher H2 concentrations (20 and 50%H2). Single-layer graphene (SLG) is obtained for Cu annealed at low H2 concentration (0 and 2.5%H2), as shown in Fig. 8(a). Graphene on such Cu surface could not have a uniform coverage (but rather a partial coverage) due to the presence of high surface features, e.g., Cu-stepped structure, hillocks, and grain boundaries. For such specimens, the corrosion species follow the straight path through the graphene and attack the surface of Cu. On the other hand, multilayer (4-5 layers) graphene is obtained when Cu is annealed under 20 and 50%H2 during the annealing process. The corrosion species follow the zigzag path through multilayer graphene (MLG) domains when subjected to corrosion electrolyte. The corrosion species take more time than the former specimens to reach the surface of Cu (Fig. 8b). Thus, the corrosion rate is slower, implyingan improved corrosion performance of Cu for MLG than the SLG specimens.

Fig. 8
figure 8

Schematic representation of corrosion species (O2 and H2O) following (a) a straight path through single/bilayer graphene (SLG) domains; (b) and tortuous path through multilayer graphene (MLG) domains. Cu exhibits rough surface when annealed under zero/low H2 while smooth surface if annealed at high H2 concentrations (20 and 50%H2)

Full size image

Conclusion

In this work, we deposited graphene films of different thicknesses (bilayer and multilayer) on Cu substrate and investigated the corrosion performance of graphene/Cu in chloride ions solution, 0.5 M sodium chloride, a concentration of salt close to that of sea water. The graphene was grown by altering hydrogen (H2) concentration during the annealing stage using CVD technique. The surface morphological features of graphene/Cu were examined by SEM and AFM. SEM and AFM results clearly highlighted the variation of graphene thickness and roughness of Cu surface obtained under different H2 concentration. At 20 and 50% H2 concentrations, 4-5 layers of graphene were obtained as confirmed by Raman spectroscopy and TEM analysis. These graphene films offered the best corrosion protection against electrochemical degradation. By Tafel analysis, it was found that the corrosion rate of G-Ann-20%H2 specimen was nearly three orders of magnitude lower than that of annealed Cu. The results obtained in this study suggest that corrosion properties of graphene/metal can be improved by optimizing CVD conditions and tailored for specific applications where resistance to salt is an important factor.