Gas Desorption and Secondary Electron Emission from Graphene Coated Copper Due to E-Beam Stimulation

Abstract

The gas desorption and secondary electron multiplication induced by electron bombardment tend to induce severe low-pressure discharge effects in space microwave device cavities. Nevertheless, few studies have focused on both secondary electron emission and electron-stimulated gas desorption (ESD). Although the suppression of secondary electrons by graphene was found to be better in our previous study, it is still unclear whether the surface modification of graphene, which brings about different interfacial states, can also be manifested in terms of ESD. The deep mechanism of gas desorption and secondary electron emission from this extremely thin two-dimensional material under electron bombardment still needs further investigation. Therefore, this paper investigates the mechanism of graphene modification on Cu metal surface on the gas release and secondary electron emission properties under electron bombardment. The surface states of graphene-modified Cu were characterized, and the ESD yield and secondary electron yield of Cu/GoCu were investigated using a self-researched platform and analyzed using molecular dynamics simulations and electron Monte Carlo simulations. The results of the study showed that the most released component on the Cu surface under the bombardment of electrons was H₂O molecules, while the most released component on the GoCu surface was H₂ molecules. The graphene-modified samples showed a significant suppression effect on the secondary electron yield and ESD only in the low-energy region below 400 eV. This study can provide a valuable reference for suppressing low-pressure discharge and multipactor phenomena in space microwave components.

1. Introduction

The low-pressure discharge effect of spacecraft microwave devices at high power in space is an essential factor threatening spacecraft loads [1,2]. Space microwave payloads will develop in the direction of high power and high reliability [3,4]. However, the low-pressure discharge phenomenon caused by the resonant multiplication of electrons in a low-pressure environment can lead to higher noise levels, lower output power, increased VSWR (Voltage Standing Wave Ratio) of the microwave transmission system, increased reflected power, channel blockage, and other problems [5,6]. Improving the threshold of low-pressure discharge to achieve high-power tolerance is necessary for all microwave devices to be in a ready condition [7,8]. In addition to the design of the device structure to avoid an excessive concentration of local microwave fields, the most important means is to improve the discharge threshold by surface modification methods [9,10]. The critical point focuses on electron-stimulated gas desorption and secondary electron emission at the surface, because of the important role they play in the discharge process.

In the space microwave device cavity, the seed electrons will resonate under the action of a high-power microwave field. When there is a low-pressure environment in the cavity (usually 10⁻³–10 Torr), the resonantly moving electrons will collide with the gas molecules by ionization, excite more electrons, and undergo the resonance multiplication process [7,11]. In this process, some electrons usually collide with the material interface to excite secondary electrons and the attached gas, thus greatly facilitating the discharge process [12]. In fact, even if the gas molecules in the cavity are extremely low (pressure below 10⁻⁵ Torr), the cascade multiplication process induced by secondary electron emission alone can lead to multipactor phenomena under the action of high-power microwave fields [13,14]. Moreover, in the multipactor process, the desorption of gas induced by electron bombardment and power-thermal effects also tends to cause more severe low-pressure discharges [15,16]. From the microphysical perspective, when electrons bombard the material surface, elastic and inelastic scattering processes occur with the atoms inside the material [17]. The energy loss accompanying the inelastic scattering process is transferred to the outer layer electrons of the atoms and excited to form the inner secondary electrons [18,19]. When the electrons on the surface cross the interface potential barrier, it forms the emitted secondary electrons [20,21]. Similarly, when electrons interact with adsorbed molecules on the material surface, the transferred energy helps the adsorbed molecules overcome the surface binding energy with direct excitation or de-excitation to generate free gas molecules [22,23].

Current methods for suppressing secondary electron emission from surfaces include surface conformation and coating modifications [24,25,26]. The surface configuration approach usually achieves a reduction in secondary electron yield by constructing a surface trap structure and using the shielding effect of the microstructure to limit the emission of secondary electrons [27]. The coating modification method reduces the secondary electron production and emission by covering the surface of the target material with a layer of low secondary electron emission [28,29]. This coating method to suppress secondary electron emission is also applied to electron-induced gas desorption, such as TiN coating [30,31,32]. Nevertheless, relatively few studies have focused on both secondary electron emission (SEE) and electron-stimulated desorption (ESD). ESD originates from the action of electrons on gas molecules [33], and in addition to the action of incident electrons, the emission of secondary electrons may also contribute to ESD. Graphene is widely used in electronic devices due to its excellent electrical properties, extremely high surface conductivity, and high thermal conductivity [34,35]. In our previous study, we also found that graphene has a good performance in suppressing secondary electrons [36,37]. However, in addition to the contribution of secondary electron emission, the microwave resonance-induced discharge is also closely related to the ionization excitation of electrons–gas collision [38]. When the number of electrons increases dramatically, ESD becomes the main source of gas, which is a critical indicator of the transformation of the multipactor into a low-pressure discharge process [12,33]. Then, the possibility of using graphene coating to suppress both SEE and ESD becomes crucial. It should be noted that surface modification of graphene changes the interfacial state [39], including the surface micromorphology, surface adsorption state, graphene layer number, coverage, and even interfacial crystal orientation, all of which may affect the surface ESD process [40,41,42]. Therefore, whether this suppression effect of graphene can also be demonstrated in ESD still needs to be investigated. The mechanism of gas desorption and secondary electron emission from the two-dimensional material under electron bombardment still needs further investigation. Therefore, the present study focuses on electron emission and the gas desorption properties of graphene-modified metallic material surfaces.

In this paper, ESD and SEE characteristics of the graphene modification on the Cu surface are investigated. Firstly, the graphene surface obtained by the chemical vapor deposition technique is characterized to clarify the surface coverage and the number of layers, surface morphology, surface state, and the crystallographic orientation characteristics of the base material. Then, a self-built vacuum test system is used to compare and test the effect law of graphene modification on the gas desorption characteristics and secondary electron emission yield under different energy electron bombardments. Finally, we used molecular dynamics simulations to analyze the stable adsorption configurations and adsorption binding energies of several major constituent gas molecules on the surface with and without graphene modification. In order to find the mechanism of deep interaction under electron bombardment, we also analyzed the scattering energy loss characteristics under the electron bombardment of graphene-modified materials using Monte Carlo simulations. The results of this study can provide a valuable reference for the suppression of low-pressure discharge and multipactor phenomena in space microwave components and also have important implications for the suppression of the surface-resolved adhesion mechanism of graphene.

2. Experiment Methods and Results

The Copper foil covered with graphene by Chemical vapor deposition (CVD, graphene coverage ≥99%, Chongqing Graphene Technology Co., Ltd., Chongqing, China) and pristine copper foil without graphene purchased from the same company were used in this study. The thickness of foil was about 25um. Surface morphology features were systematically analyzed on the different scales by various technologies, including light microscope (Olympus BX-51, Tokyo, Japan), scanning electron microscopy (SEM, TM3030Plus, Hitachi, Tokyo, Japan), and atomic force microscope (AFM, Dimension ICON, NanoScope V, Bruker, Bremen, Germany). The element components and chemical states were examined by X-ray photoelectron spectroscopy (XPS, Thermo Fisher ESCALAB, Racine, WI, USA. The Raman spectra were also collected at room temperature with a backscattering configuration by Raman spectrometer (Renishaw InVia Qontor, Kingswood, UK) (532nm, 100 mW). The crystal structure was analyzed by X-ray diffractometer (XRD, Rigaku Smartlab with Cu Kα X-ray source, Tokyo, Japan).

2.1. Surface States Measurement

2.1.1. The Multiscale Morphology and Crystal Structure of Cu Substrate

The light microscope was first used to analyze the surface characteristics on a relatively large scale. As shown in Figure 1, the pristine copper foil reveals a relative wrinkled and rough surface without obvious grains. In contrast, well-developed large grains with sizes between 50 µm to 200 µm are clearly observed on the surface of Copper foil covered with graphene. This indicates that the heating treatment at high temperatures during the CVD process promotes crystallization.

Since light microscopy is not able to detect sub-micrometer scale structures due to the limitation of visible light wavelength, scanning electron microscopy was used to investigate the possible delicate structures. As shown in Figure 2a,b, the pristine copper foil only shows a wrinkled and rough surface, consistent with polarizing microscope images. This is believed to be characteristic of the rolling process during the preparation of copper foil. Besides the clear observation of large grains, several distinct features were found in the copper foil covered with graphene. As shown in Figure 2c, different grains and even different parts within one single large grain always display distinct contrast, which implies diverse crystallography orientations. As shown more clearly in Figure 2d, the straight boundaries within one grain separating regions with different contrast are believed to be domain walls. In addition, fine wrinkled structures are found to be widely distributed in the scanning region.

In order to further investigate the details of fine wrinkled structures, the atomic force microscope probing was carried out on the surface of copper foil with graphene in a region of 3 µm × 3 µm. As shown in Figure 3, the Zig-Zig-type wrinkled structures are confirmed in three dimensions. The cross-section profile of the selected line is plotted in the right panel of Figure 3. The wrinkle is about 220 nm in width and 10 nm in depth. The calculated bending angle is below 3 degrees. An overall tilt on the sample surface is observable from Figure 3a but it should not affect the obtained main features of rinkled structures, considering the fact that the tilt occurs in a much larger scale than the fine structures. No noticeable signal variations related to graphene can be seen in the scanning region, suggesting good uniformity and completeness.

The room temperature XRD measurements were used to investigate the crystal structure of copper foils. As shown in Figure 4a,b, both copper foil samples demonstrate characteristic diffraction peaks related to Cu single phase. The main peaks are indexed in the patterns. Note the (220) diffraction peak is dominating in pristine copper foil while the (311) peak becomes prominent in the graphene-modified copper sample. Considering the large grain size in the graphene-modified Cu sample, the significantly increased signal intensity on one peak likely represents the orientation of a given large grain under detection. In addition, the diffraction peaks of graphene-modified Cu foil are much sharper than those of original Cu foil, as evidenced in the inset of the Figure 4b. The crystalline grain size can be calculated by the Scherrer formula $\mathrm{D}=\frac{\mathrm{K}\mathsf{\lambda }}{\mathrm{B}\cos {\mathsf{\theta }}^{\prime }}$ (D is grain size, K is Scherrer constant, λ is X-ray wavelength, θ is the diffraction peak position, B is the half width of the diffraction peak). It is evident in the formula that the sharper the diffraction peak, the larger the grain size under the same measurement conditions. The calculation results suggest the grain size of ~150 nm for the pristine copper foil and 100 µm for copper foil with graphene, respectively, which is generally consistent with the abovementioned microscopy observations.

2.1.2. The Raman Investigations

Raman spectroscopy is a powerful and widely used method to analyze graphene, which enables one to obtain information about the number of graphene layers and the quality/defect states of graphene [43,44]. Generally, carbon-based materials have three main Raman characteristic peaks, namely G peak near 1580 cm⁻¹, D peak near 1350 cm⁻¹, and G′ (2D) peak near 2700 cm⁻¹. The G peak is caused by the E₂g pattern in the center of the first Brillouin region, and the D peak is always related to defect and impurity states. As shown in Figure 5, the Raman spectrum of graphene on Cu foil is collected in a reflection mode at room temperature. It can be seen from the figure that there are two obvious characteristic peaks of graphene, the G peak at about 1580 cm⁻¹ and the G′ (2D) peak at about 2680 cm⁻¹. The ratio of peak strength of peak G to peak 2D is 0.4, indicating that the graphene has about 1–2 layers [45,46]. The half-height and width of 2D peak is 35 cm⁻¹. No obvious defect peak, e.g., the D peak at 1380 cm⁻¹, can be found within the detection limit, suggesting the defects (if any) are very rare. Overall, the Raman investigations strongly indicate that the copper foil is modified by the monolayer/bilayer graphene with good quality, completeness, and defects-free.

2.1.3. X-ray photoelectron spectroscopy

The effects of graphene on the surface chemical states are also investigated by X-ray photoelectron spectroscopy. The results are shown in Figure 6. Obviously, the C related peaks become dominant in the graphene-modified copper foil, confirming the presence of graphene. Accordingly, the Cu signal is reduced in graphene-covered copper foil compared with the original foil. In addition, the pristine foil shows the existence of N signals, implying possible absorption of N₂ from the air. In contrast, no obvious N signals can be seen in the copper foil with graphene, suggesting the graphene can prohibit the gas absorption effects of the Cu foils.

2.2. Gas Desorption and Secondary Electron Emission

2.2.1. Electron Stimulated Desorption

In this study, the electron-induced gas desorption and secondary electron emission in situ measurement on material surfaces were performed in a self-developed high vacuum system, as shown in Figure 7. The platform consists of two vacuum chambers: the pre-vacuum chamber and the main test chamber. The ultimate vacuum of the main test chamber is 10⁻¹⁰ Torr, which is maintained by a three-stage vacuum pumping system (mechanical and molecular pumps from Pfaff (Varian, Milan, Italy) and ion pumps from Gamma Vacuum, Shakopee, MN, USA). The electron gun of the equipment is a coaxial grid-controlled electron gun of Staib Instruments (Langenbach, Germany), with an energy range of 20 eV–5000 eV and an E-beam current range of 0.05 nA–50 uA, located directly above the sample table in the analysis chamber, as shown in Figure 7. For metallic conductors, the secondary electron yield can be obtained by the incoming and outgoing currents on a sample surface [47]. For obtaining the electron currents, the system is equipped with a Keithley Instruments model 6487E picoammeter (Cleveland, OH, USA). The picoammeter connected between the sample stage and ground is used to apply bias to the sample and to detect the current flowing through the sample stage when measuring secondary electron emission characteristics. The testing current range is 2 nA–20 mA, with a bias voltage range of +/−505 V, a measurement accuracy of 10 fA, and a typical root-mean-square noise of 20 fA at a current of 2 nA. The incident E-beam current of the electron gun can be received through a Faraday cup at the front of the sample holder and then measured with the picoammeter. The emission secondary electron current can be obtained by the difference between the incident current and the leakage current at the bottom of the conductor sample.

Due to the high vacuum and low gas leakage rate of our platform, the material gas desorption characteristics induced under electron bombardment can be obtained by detecting the change of air pressure inside the chamber under electron bombardment. Furthermore, the quadrupole mass spectrometer (QMS) (Pfeiffer, Aßlar, Germany ) is used for the residual gas analysis (RGA) in the vacuum chamber.

The gas desorption characteristics of graphene surface modified copper and unmodified copper under electron irradiation are tested as follows. The sample to be tested is placed on the tray. The vacuum level of the chamber is ensured to be below 10⁻⁸ torr by continuous evacuation with a three-stage vacuum pump. Then, all vacuum valves in the chamber are closed. The sample is irradiated with a specified energy electron beam for 200 s. At the same time, the vacuum level in the chamber is monitored by a vacuum gauge to obtain the dynamic rate of change of air pressure for electron radiation-induced gas desorption. The gas composition of the chamber is measured using the RGA in the case of chamber air pressure greater than 10⁻⁷ Torr and less than 10⁻⁵ Torr.

The results are shown in Figure 8a,b, with the vertical coordinate being the air pressure inside the main chamber (corresponding to the number of molecules in the chamber). The barometric pressure increases approximately linearly and monotonically with increasing irradiation time. Compared with the pressure increase rate of the background case and the Cu/GoCu loading case without electron irradiation, the electron bombardment-induced gas desorption is much larger than the system background leakage and the gas release from the sample itself. Therefore, it can further illustrate the method’s reliability in testing the electron-induced gas desorption characteristics. At the same time, the linear increase in air pressure (or gas release) with time indicates that the degree of gas desorption due to electron bombardment does not significantly change the surface adsorption state.

According to Ideal Gas Law, we can calculate the electron-stimulated desorption rate γ as shown in Figure 8c. The electron-stimulated gas desorption also needs to be removed from the system gas leakage of the sample to be tested by loading. The vertical gas desorption rate in Figure 8c represents the number of gas molecules stimulated by one bombarding electron. From the figure, we can see that although high gas desorption rates for both Cu and GoCu correspond to the high-energy electron irradiation case, the gas release rate does not necessarily increase monotonically with the bombarding electron energy in the low-energy case. For the unmodified Cu surface, γ tends to decrease and then increase in the low-energy region, reaching a minimum at an electron energy of 200 eV. However, for the graphene-modified Cu surface, γ still shows a consistent trend of monotonically increasing, despite the change in the growth slope at 200 eV. In the low-energy region below 400 eV, the graphene-modified samples show a significant gas release suppression effect. In contrast, the suppression effect gradually decreases in the higher energy region and is even slightly higher than that of the comparison unmodified samples.

The reason for the non-monotonic variation of the desorption rate in the case of low-energy electron irradiation may be related to the secondary electron emission from the surface, which we will discuss and analyze later in conjunction with the secondary electron and electron-material interaction processes.

In order to further investigate the effect of electron beam irradiation on the surface state, the evolution of the released gas composition induced by electron bombardment at different energies with and without surface graphene modification needs to be further investigated. Here, the desorption gas composition is analyzed by an in situ residual gas analyzer RGA in the platform, which also requires removing the background gas components. Figure 9 shows the results of ESD gas components for those two different surface cases. The dominant gas types include H₂O, H₂, N₂, and CO₂. In the absence of electron irradiation, the contents in the chamber are H₂O > H₂ > N₂ > CO₂. It should be noted that the ESD gas components in the figure are removed from the effect of intrinsic gas release. Under electron bombardment, there is a clear difference between the components of gas release from surface graphene-modified Cu and unmodified Cu surfaces, with the most released H₂O molecules from Cu surfaces and the most released component from GoCu surfaces being H₂ molecules. For the graphene-modified surface, the change of the ESD gas components was weak when the bombarding electron energy was below 200 eV. The difference of the ESD gas components was further enlarged only when the energy was further increased to larger than 300 eV. This situation may be related to electrons’ penetration ability and gas release from the interfacial layer, which we will discuss later. In addition, for the low-component gas ions CO₂ and OH⁻, the electron bombardment on different surfaces shows slightly different trends.

2.2.2. Secondary Electron Emission

The secondary electron emission induced by electron bombardment is an important process in the microwave high-power discharge process of spacecraft, and the emission of secondary electrons may affect surface gas desorption as well. Therefore, to further investigate the effect of electron beam irradiation on the surface state, the secondary electron yields (SEY) are investigated under different irradiation conditions, and the results are shown in Figure 10. The SEY increases rapidly with the increase in irradiated electron energy and reaches a maximum, and then starts to decrease with a further increase in electron energy. The maximum secondary electron yield δmax of the graphene-modified Cu sample is 1.34 lower than that of the unmodified Cu at 1.49. The electron irradiation energy corresponding to the maximum yield E_max increases from 200 eV to 300–400 eV. However, it should be noted that when the incident electron energy is higher than 400 eV, the suppressive effect of the graphene surface modification on the SEY of the Cu surface disappears, and the SEY on the GoCu surface is even greater than that on the Cu surface. At this time, the SE emission from the surface is promoted. The opposite effect of graphene modification on the suppression of the Cu surface SEY in the high-energy region may result from the combined effect of the internal electron scattering cross section and the surface work function, as we will discuss subsequently.

3. Discussion

The ESD process requires the acquisition of energy to help the surface-adsorbed gas molecules overcome the surface binding energy, including direct collision energy transfer and de-excitation energy transfer modes. Therefore, the adsorption binding energy of gas molecules to the material surface is the primary energy barrier for the desorption of gas molecules and is directly related to the ESD situation. Hence, we calculated the adsorption binding energy of the dominated gas molecules with and without graphene-modified surfaces by employing a molecular dynamics approach. In the simulations, we chose the COMPASS force field, and the convergence of the structural optimization is set as follows: Energy: 2 × 10⁻⁵ kcal/mol, Force: 10⁻³ kcal/mol/A; Displacement: 10⁻⁵ A. In order to obtain the lowest energy stable configuration, the Quench simulation is performed for the adsorption system at 300 K temperature. The structure optimization is performed every 500 steps to obtain the 100 lowest energy state configurations, and the lowest energy state is selected to obtain the adsorption binding energy.

Based on the results of surface XRD patterns in Figure 4, we select the two most dominant crystal orientation cases, (220) and (311), in the molecular structure simulation and perform interface matching by Redefine Lattice and Supercell methods. According to the ESD gas components in Figure 9, in the simulation, we choose the two most abundant gases, H₂O and H₂, for the binding energy calculation. One of these two gas molecules is a polar molecule, and the other is a nonpolar molecule.

The results of the adsorption binding energy of these two gas molecules on four different surfaces are shown in

Table 1. Molecular adsorption binding energy of two typical gas molecules at different crystal directions and modified state interfaces.

Molecular Adsorption Energy (kcal/mol)	H2	H2O
Cu (311)	−3.84	−7.80
Graphene on Cu (311)	−0.14	−3.23
Cu (220)	−1.08	−3.80
Graphene on Cu (220)	−0.16	−2.28

. For example, the adsorption binding energy of H₂ on Graphene on Cu (311) is only −0.14 kcal/mol, while the adsorption binding energy of H₂O on this surface is −3.23 kcal/mol. In addition, the modification of graphene also significantly reduces the adsorption energy of gas molecules on the surface. To further explain the binding energy effect, Figure 11 shows the interfacial configuration of the gas molecules for four typical adsorption cases. Compared with copper in the (220) crystal orientation, there is considerable oblique embedding space on the (311) surface (Figure 11a). The adsorption binding energy of H₂ and H₂O on the (311) surface (−3.84 kcal/mol and −7.80 kcal/mol) is significantly larger than that of the (220) surface case (−1.08 and −3.80 kcal/mol).

The graphene layer increases the spacing between the adsorbed gas molecules and the substrate, and the intense bonding energy of C−C/C=C in graphene plays a particular isolation role, which makes the adsorbed molecules interact with the surface less and the adsorption binding energy weak. For the nonpolar gas molecule H₂, the stable adsorption configuration of the surface is basically the same for both graphene-modified Cu(311) and Cu(220) surfaces: parallel to the interfacial layer and at the center of the graphene ring (Figure 11b). Thus, H₂ presents similar adsorption binding energies (−0.14 kcal/mol and −0.16kcal/mol) on the graphene-modified surface. As for the polar molecule H₂O, the graphene-modified Cu(311) and Cu(220) surfaces still show different adsorption stabilization configurations: the stabilization configuration of H₂O at the (311) interface is tilted. In comparison, the stabilization configuration of H₂O at the (220) interface is parallel to the interface (Figure 11d), corresponding to binding energies of (−3.23 kcal/mol and −2.28 kcal/mol). It indicates that the nonpolar H₂O molecules are still subjected to the lower Cu atom array, although the surface binding energy is weakened by the surface modification of graphene.

The more extensive adsorption binding energy corresponds to a more stable surface adsorption state. Thus, the electron-stimulated gas desorption rate of H₂O gas is more significant on the Cu surface without graphene modification than H₂ (e.g., Figure 9). For the graphene-modified Cu surface, the weakened binding energy leads to a weaker desorption of H₂O gas. The same weakened binding energy of H₂ molecules is enhanced instead, probably due to the presence of H₂⁺ in the free or suspended state at the interface between graphene and Cu, where small molecules of H₂ can desorb outward through the graphene layer under the electron bombardment.

From Figure 8 and Figure 10, we can see that both the ESD rate and the SEY are related to the incident electron energy. Hence, we further analyze the desorption phenomenon of the Cu surface with graphene modification by simulating the electron-surface interaction process. The electron-material interaction includes elastic scattering and inelastic scattering. The Monte Carlo simulation method is used here for the electron–material interaction. The Mott scattering model is used for describing elastic scattering, and the Joy and Luo modified continuous energy loss model is used for describing inelastic scattering. The material model is set to be a graphene layer of one atomic layer thickness (0.34 nm) covering the surface of the Cu substrate.

Figure 12 shows the energy loss distribution of electrons inside the material after electrons of four different energies (50 eV, 100 eV, 200 eV, and 400 eV) are incident on the graphene-modified Cu surface. Electrons of 50 eV and 100 eV energies mainly interact with the surface graphene layer, and 90% of the energy loss is gathered within 0.8 nm of the surface. When the electron energy is increased to 400 eV, the 90% energy loss range increases to 3.4 nm, ten times the thickness of the graphene layer. Due to the differences in material scattering cross sections, energy loss functions, and potential interfacial barriers, the energy loss distribution of electrons can exhibit discontinuities at the Graphen–Cu interface. Although the total average free range of electrons after considering phonon interaction is smaller in the low-energy region, the several nm average free range still ensures that a large number of Cu-excited internal secondary electrons exit from the surface at a high-energy electron incidence (e.g., 400 eV). In Figure 10, the graphene modification suppresses secondary electrons at incident electron energies below 400 eV, but the corresponding secondary electron yields are higher at higher energies. It is mainly because, for Cu, graphene suppresses secondary electron emission mainly by taking advantage of the low scattering cross section of the C atom to introduce the incident electrons into the interior of the material, thus reducing the number of electrons emitted from the surface. However, when the energy of incident electrons increases, many electrons interact with Cu in the inner layer. The outgoing secondary electrons come from the excitation of Cu in the inner layer, which significantly weakens the inhibitory effect of surface graphene. More importantly, just as graphene coatings can reduce the surface work function of some metals to enhance the field emission [48], for Cu, the modification of the graphene layer can also reduce the work function of the surface layer, which is a parameter closely related to the secondary electron emission. A lower work function weakens the potential barrier for internal electrons to exit from the surface and promotes the emission of secondary electrons.

By comparing Figure 8c and Figure 10, it can be found that there is a strong correlation between the suppression of ESD and the suppression of SEY by graphene modification, and the suppression effect of graphene modification on both of them disappears at 300–400 eV and is enhanced in reverse after that. This indicates that the secondary electrons also interact with the adsorbed gas molecules on the surface during the outgoing process, and part of the contribution of the electron bombardment-induced gas release rate comes from the outgoing secondary electrons. In addition, for graphene-modified Cu, the ESD rate should also have a contribution from the molecular excitation at the layers interface.

4. Conclusions

This paper investigates the mechanism of the modification of graphene on the surface of metallic Cu on the gas release properties and secondary electron emission properties under electron bombardment. The high-temperature heating treatment during the growth of graphene by CVD using chemical vapor deposition promotes the formation of crystallites, which changes the original Cu whole foil wrinkled surface into large particles ranging from 50 μm to 200 μm in size. It causes the Cu crystal orientation to change from (220) dominant to (311). The copper foil was modified by monolayer/bilayer graphene, intact, defect-free, with the presence of a Zig-Zig type wrinkled structure with a width of about 220 nm, as characterized by Raman spectroscopy. Using the self-built gas release and secondary electron research platform for testing, we found that the most released molecules from the Cu surface were H₂O, while the most released component from the GoCu surface was H₂ molecules. Through molecular dynamics simulations, it was found that although the nonpolar H₂O molecule was subjected to surface modification of graphene to weaken the surface binding energy, the molecular conformation was still influenced by the array of underlying Cu atoms. For the unmodified Cu surface, the ESD rate decreased and then increased in the low-energy region, and the ESD rate reached its lowest when the electron energy was 200 eV. The graphene-modified sample only shows a significant gas release inhibition effect in the low-energy region below 400 eV. The graphene modification reduces the maximum secondary electron yield δmax of Cu samples from 1.49 to 1.34. When the energy is further increased beyond 400 eV, the suppression effect of graphene on the SEY of Cu disappears due to the combined effect of the scattering cross section and the surface work function as the incident electrons penetrate deeper. The results of this study can provide valuable references for the suppression of low-pressure discharge and multipactor phenomena in space microwave components and also have important implications for the suppression of surface desorption mechanisms by graphene.