With the growing requirement of high frequency electronics and devices, Terahertz (THz) technology [1] has been found to have great potential in a wide range of applications such as biological sensing [2], spectroscopy and imaging [3], wireless communication [4], and detection and identification [5]. THz radiation usually contains electromagnetic waves in the frequency of 0.1-10 THz [6], 1-4 orders of magnitude higher than the frequency of microwave. This means that the THz wave has larger available transmission bandwidth, as well as the ability to carry more information [7]. Recently, Federal Communications Commission (FCC) decided to open up “terahertz wave” as experimental spectrum for 6 G services. However, the rapid development of THz science and technology is bound to generate a lot of terahertz waves, which will inevitably cause electromagnetic interference (EMI) among the electronic elements inside the electronics. Therefore, EMI shielding in the corresponding frequency range is urgently needed. So far most of the studies on the shielding properties of different materials focus on the frequency range from MHz to GHz [[8], [9], [10]], while the studies in THz band are relatively few, which is worth paying attention to.

For high-performance EMI shielding, materials with high conductivity, minimal thickness, light weight, good flexibility and high stability are highly desirable, especially for the applications in aircraft, aerospace and flexible electronics such as portable electronics and wearable devices [11]. Some promising progress has been made in this regard. Cataldi et al. fabricated cellulosic graphene biocomposites with the thickness of only ≈70 μm, and the average EMI shielding effectiveness (SE) can reach 40 dB at 0.5-0.75 THz [12]. But the resistance of the film increased to 1.8 times after 40 cycles of folding-unfolding, which usually cause the decrease of SE. Dong et al. prepared flexible reduced graphene oxide papers by solution evaporation method, which exhibited an EMI SE of 72.1 dB at 0.6 THz with the thickness of ≈370 μm [13]. Choi et al. prepared MXene/nano-metamaterial by drop-casting method, which demonstrated the EMI SE of 20 dB with a 150 nm thick Mxene film [14]. However, its flexibility was not considered, and the SE was relatively low. Up to now, most existing ultrathin THz shielding materials have their thickness above micron-size thickness, and the remaining a few THz shielding materials with nanoscale thickness behave very mediocre shielding performance. Therefore, it is still a big challenge to fabricate THz shielding materials with excellent comprehensive performance including high EMI SE, minimal thickness, good flexibility and stability.

For high-frequency electromagnetic wave, metals with high conductivity are effective shielding material, and Cu has been commonly used because of its low price and easy availability. But it is easy to be oxidized in air, especially for Cu with micro/nano scale thickness and in high temperature environment of electronic devices, which can cause obvious decrease of its conductivity and further adverse effect on its shielding performance. And Cu can be damaged due to multiple bending when used for EMI shielding in flexible electronic devices, which will lead to a reduction in the shielding effect. Graphene is an ultrathin two-dimensional carbon material with high strength, modulus and impermeability, and superb thermal and chemical stability. Therefore, here we prepared copper/graphene (Cu/Gr) nanolayered composites by alternating deposition of Cu and wet-transfer of graphene on substrates. Cu film can achieve a very high EMI SE with a very small thickness, and the additional graphene layer can significantly improve the resistance to oxidation and fatigue, with just a very small content in the composite. The average SE value of the obtained composite reached 44.17 dB and 60.95 dB with the thicknesses of only 40 nm and 160 nm respectively at 0.1-1 THz. Moreover, due to the excellent impermeability and superb thermal and chemical stability, graphene in composites can effectively protect Cu from oxidation [15]. After heating at 120 °C for 3 h in air, the maintenance rate of EMI SE of the composite was still able to reach 93.09%, while that of pure Cu film was only 62.15%. Besides, the addition of graphene improved the flexural fatigue resistance because of the ability of graphene to block and deflect crack propagations and suppress the formation of fatigue cracks [16]. After 1500 bending cycles, the EMI SE value of the composite showed a maintenance rate of 98.87%, obviously higher than that of multilayer Cu film (93.07%). Importantly, the THz shielding properties of metal/graphene composites have not been reported previously. Here we firstly demonstrated the construction of ultrathin flexible terahertz shielding materials based on Cu/Graphene nanolayered structure. It has not only high EMI SE and ultrathin thickness, but also excellent stability and EMI SE reliability, which may exploit the opportunity for their application in next generation communication and electronics.

Graphene was prepared by chemical vapor deposition (CVD) method. The commercial Cu foil with thickness of 30 μm was cleaned with deionized water, acetone and ethanol in turn. Then the cleaned Cu foil was treated with FeCl₃/HCl etching solution, washed with deionized water, and blown dry with nitrogen. Then, the treated Cu foil was loaded into a tube furnace (Hefei Kejing Company OTF-1250X-80-SL) to grow graphene as follows. Firstly, the tube furnace was heated to 1000 °C at the rate of 10 °C/min under the protection of Ar (500 sccm). Then the Cu foil was annealed at 1000 °C for 40 min with additional H₂ (20 sccm), and CH₄ (6 sccm) was introduced as the carbon source for graphene growth for another 13 min. Afterwards, the system was cooled down to room temperature naturally with Ar (500 sccm) and H₂ (20 sccm). Finally, the Cu foil on which graphene grew on was taken out of the furnace for subsequent use.

The Cu/Gr nanolayered composites were prepared by alternating deposition of Cu thin film and wet-transfer of graphene on substrates. Firstly, Cu thin film was deposited on a PI film by vacuum evaporation. Then the polymethyl methacrylate (PMMA)/anisole solution was spin-coated onto the prepared graphene on Cu foil to form a support layer, and the Cu foil was etched by CuSO₄/HCl solution. Then, the obtained graphene/PMMA film was transferred onto the Cu thin film deposited on the PI substrate, and the PMMA layer was removed by acetone subsequently. The deposition of Cu thin film and the transfer of graphene were carried out alternately to form the Cu/Gr nanolayered composites (the thickness of each layer of Cu thin film was set to be 40 nm). Besides, multilayer Cu films without graphene were also prepared by the same process.

The morphology and energy dispersive spectroscopy (EDS) of the Cu/Gr nanolayered composites and multilayer Cu films were investigated with a scanning electron microscope (SEM; JSM-7800, Japan). The Raman spectra were investigated using a SR-500I-A Raman spectrometer with excited laser at 532 nm. The optical images were acquired using an upright metallurgical microscope (Leica DM750 M). The X-ray diffraction (XRD) patterns were obtained by a Rigaku Smart Lab 3 kW diffractometer. The sheet resistance was tested using a digital and intelligent four-probe meter (ST2258C). X-ray photoelectron spectroscopy (XPS) results were obtained by an ESCALAB 250Xi system from Thermo Scientific. For the convenience of test and avoiding the interference of substrate, samples tested by SEM, Raman spectrometer and XPS were prepared on silicon wafer substrates. All measurements were conducted at room temperature (≈25 °C).

The terahertz shielding effectiveness of samples was measured by the THz time-domain spectroscopy (THz-TDS) system at room temperature. The transmitter and the receiver are commercial photoconductive antennas, and the laser is a femtosecond fiber laser with a central wavelength of 800 nm, and a pulse width of 75 fs. The samples were attached to a hollow aluminum plate for test. Terahertz wave focused on the sample with a spot size of 2.5 mm. The EMI SE of samples is defined in decibels (dB) as follows [17]:

where P_I (E_I) and P_T (E_T) denote the incident power (field strength) and transmitted power (field strength) of electromagnetic waves respectively.

Fig. 1 shows the fabrication process of the Cu/Gr nanolayered composites. The graphene was prepared by CVD method, and then the composites were prepared by alternating deposition of Cu by vacuum evaporation and wet-transfer of graphene on substrates. The prepared samples with the increasing of the number of layers were denoted as 1Cu, 1Cu1Gr, 2Cu1Gr, 2Cu2Gr, etc. The cross-sectional SEM images of the 4Cu4Gr sample on silicon wafer are shown in Fig. 2(a) and (c), and that of 4Cu sample is shown in Fig. 2(b). It can be seen that the 4Cu4Gr sample exhibits obvious layered structure (especially in Fig. 2(c)), while the 4Cu sample do not show a clear layered interface. Van der Waals interaction exists at the interface between Cu and graphene, as known from density functional theory (DFT) calculation results [18]. Besides, from the scale on the image, it can be known that the thickness of the 4Cu4Gr sample is 158.33 nm. Therefore, the thickness of one layer of Cu with graphene in the Cu/Gr nanolayered composites discussed below is considered to be about 40 nm, which is consistent with the thickness (40 nm) set in the process of Cu deposition. In order to observe the surface morphology of the composite, a layer of graphene was transferred onto a layer of Cu deposited on a silicon wafer. Fig. 2(d) shows the area of the Cu layer covered by the graphene layer completely. The Cu particles under the graphene layer are faintly visible due to the covering of graphene. For the convenience of observation and comparison, the area of the Cu layer covered with the graphene layer partially was selected specially, as shown in Fig. 2(e) and (f). The boundary of the graphene layer transferred onto the Cu layer is marked with a white dotted line. The dark part (upper left) is the area covered with graphene, and the light part (lower right) is the bare Cu, which is uniformly distributed in a particle state with a particle size of dozens of nanometers.

Fig. 2. Cross-sectional SEM images of (a, c) 4Cu4Gr sample with layered structure and (b) 4Cu sample without clear layered interface. Top-view SEM images of (d) the area of the Cu layer covered by the graphene layer completely and (e, f) the area of the Cu layer covered with the graphene layer partially. The boundary of the graphene layer transferred onto the Cu layer was marked with a white dotted line.Fig. 3(a) and (b) shows the Raman spectra of Cu foil after the growth of graphene and Cu/Gr nanolayered composites with different layer numbers during preparation, respectively. The obvious characteristic peaks of graphene, G peak (at about 1580 cm^-1) and 2D peak (at about 2697 cm^-1) in Fig. 3(a) indicate the successful synthesis of graphene on the Cu surface. And the extremely weak D peak (attributed to the presence of structural defects in graphene) at about 1350 cm^-1 means only minimal defect in the carbon crystalline texture, indicating the high quality of the grown graphene. Besides, the coverage rate of the graphene grown on the Cu foil can be observed from the optical images in Fig. S1 (in Supporting Information), which demonstrates the complete coverage. In Fig. 3(b), it can be seen that there is no obvious sharp peak in the Raman spectrum when the top surface is Cu layer, but when the top surface is graphene layer, the Raman spectrum shows obvious characteristic G peak (at about 1580 cm^-1) and 2D peak (at about 2685 cm^-1) of graphene. The alternating layered structure of Cu/Gr nanolayered composites can also be confirmed from this result. Moreover, despite a slight increase due to the transfer process, the D peaks of samples with different layer numbers are still very weak, indicating that the defects of graphene in the prepared composites are kept in a very low level. The red-shift of 2D peak after the transfer may be related to the strain during graphene growth on Cu foil surface [19]. Fig. 3(c) shows the XRD pattern of the 4Cu4Gr sample. It exhibits peaks at the diffraction angle of 43.3° and 50.4°, which are indexed as (111) and (200) planes of Cu with cubic symmetry respectively, without any obvious peaks of copper oxide (CuO) or cuprous oxide (Cu₂O). And the strong peak at 2θ = 43.3° indicates that the Cu deposited by vacuum evaporation possesses a highly (111) preferred orientation [20]. Fig. 3(d) shows the photographs of the 4Cu4Gr sample bended by a tweezer and released naturally. The sample can be bended easily, and there is no break or crack when it is bended and released, demonstrating the good mechanical flexibility of the Cu/Gr nanolayered composite.

The THz properties of samples were measured by a self-built four parabolic reflector THz-TDS system as shown in schematic diagram (Fig. 4(a)), and the details and optical images of the measurement system are shown in Fig. S2. Conductivity of materials is an important factor affecting the EMI SE, and the high conductivity endows the materials with high EMI shielding performance [21]. Fig. 4(b) shows the sheet resistance and conductivity of Cu/Gr nanolayered composites with different layer numbers (1Cu1Gr, 2Cu2Gr, 3Cu3Gr, 4Cu4Gr). The sheet resistance values of the composites decrease with the increase of the number of layers and the thickness. The conductivity values are calculated from the sheet resistance and thickness, which increase gradually and tend to be stable. It may be attributed to the vacancy filling caused by the repeated deposition of Cu. Besides, the conductivity values of multilayer Cu films (1-4Cu) are shown in Fig. S3. The measured SE curves of Cu/Gr nanolayered composites with different layer numbers at 0.1-1.0 THz are shown in Fig. 4(c). The average EMI SE values of samples with different layer numbers are 44.17 dB (1Cu1Gr), 54.56 (2Cu2Gr), 58.34 dB (3Cu3Gr) and 60.95 dB (4Cu4Gr), increasing with the increase of the number of layers of the samples (the SE values of 1-4Cu samples are shown in Fig. S4). The EMI SE of 40 dB means the ability to block 99.99% of the incident radiation, and 60 dB means 99.9999% of the radiation can be blocked. Therefore, the THz shielding performance of the prepared Cu/Gr nanolayered composites is excellent, which greatly exceeds the requirements of commercial EMI shielding materials (20-30 dB). Compared with the previously reported THz shielding materials [[12], [13], [14],[22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34]], the Cu/Gr nanolayered composites exhibit both high EMI SE and small thickness, and rank at the top of the comparison chart with obvious advantages as shown in Fig. 4(d). The thickness and average SE of these materials are listed in Table S1. The specific SE (SSE) is defined as the average SE divided by thickness (SE/thickness) in this paper because SE and thickness are mainly considered, and in order to compare SE and thickness of different materials comprehensively. The SSE of the Cu/Gr nanolayered composite is up to 1104.25 dB/μm (Table S1), which is an order of magnitude larger than that of the best reported THz shielding materials. These results show that the Cu/Gr nanolayered composites have great potential in the EMI shielding material for ultrathin electronics.

Cu is a kind of metal widely used in EMI shielding due to its high conductivity and low price. However, Cu is easy to be oxidized in air, especially for Cu with micro-nano scale thickness and in high temperature environmental, which can cause substantial attenuation on its shielding performance. Thus the oxidation resistance and stability of shielding materials are very important in practical applications. It has been known from the above that the 1Cu1Gr composite exhibits good THz shielding effect with the EMI SE of more than 40 dB, indicating a great potential in the electromagnetic shielding of microelectronic devices. Then the oxidation resistance of the 1Cu1Gr composite is also studied here. For the characterization of the oxidation resistance, the 1Cu1Gr and 1Cu samples prepared on silicon substrates (used for the convenience of test and avoiding the interference of substrate) were heated at 120 °C for 3 h in air. The chemical composition of the samples before and after heating were characterized using XPS and EDS.

Fig. 5(a) and (b) shows the 2p and LMM spectra of Cu, respectively. In the Cu 2p spectra, Cu(Ⅱ) species can be distinguished from Cu and Cu(Ⅰ) species, because the XPS spectra of Cu(II) ions exhibit two characteristic signals: (1) multiple line broadening of both Cu 2p_3/2 and Cu 2p_1/2 peaks, (2) intense shake-up satellite peaks at about 10 eV higher than the main Cu 2p_3/2 and Cu 2p_1/2 peaks [35]. As can be seen from Fig. 5(a), neither the above-mentioned satellite peaks nor Cu 2p peak broadening appeared in the Cu 2p spectra of the 1Cu1Gr sample before and after heating. This indicates that no Cu(II) species formed on the surface. However, there are obvious multiple line broadening and intense shake-up satellite peaks in the Cu 2p spectrum of the 1Cu sample after heating, which means the formation of Cu(II) species (such as CuO), indicating that the 1Cu sample was oxidized obviously without the protection of graphene. Cu(Ⅰ) species (such as Cu₂O) are hard to be differentiated from pure Cu in Cu 2p spectra due to the similarity of Cu 2p peaks, thus the Cu LMM spectra are obtained to discriminate between Cu and Cu(I) as shown in Fig. 5(b). The peak located around 568 eV, labeled as peak 2, represents pure Cu, and the peak around 570.3 eV, labeled as peak 3, represents the Cu(I) chemical state [36]. The ratio of I_peak2/I_peak3 can be used to identify the content of Cu or Cu₂O on the surface [35]. As shown in Fig. 5(b), the I_peak2/I_peak3 of the 1Cu1Gr sample decreased very slightly after heating, while the I_peak2/I_peak3 of the 1Cu sample decreased greatly after heating, indicating that the content of Cu₂O on the surface of the 1Cu sample increased greatly after heating in air. In addition, from EDS analysis results (only copper and oxygen were considered) in Fig. 5(c) and (d), it can be seen that the mass fraction of oxygen in the 1Cu1Gr sample changed a little (from 2% to 2.1%), while that of 1Cu sample increased obviously (from 1.9% to 9.6%). Besides, the results of Raman spectra also demonstrate a much higher degree of oxidation of the 1Cu sample than the 1Cu1Gr sample after heating at 120 °C for 3 h in air (Fig. S5). These results indicate that graphene in the Cu/Gr nanolayered composites possesses a good protective effect on Cu, and the composites exhibit much higher oxidation resistance than bare Cu. The great protective effect of the graphene layer on the Cu layer can be attributed to its excellent impermeability and superb thermal and chemical stability [15].

What’s more, the influence of the oxidation resistance on the EMI SE of the Cu/Gr nanolayered composite was also studied. For the convenience of testing, PI film was used as the substrate here (the effect of PI film on THz waves is small, which can be subtracted as background in the test), and a sandwich structure was fabricated as Gr/Cu/Gr on the substrate in order to protect the Cu film from the oxygen permeated through the PI film. Besides, the bare Cu film with the same thickness deposited on PI film was also prepared as control. The EMI SE values of the Gr/Cu/Gr sample and the Cu film before and after heating at 120 °C for 3 h were tested, as shown in Fig. 5(e) and (f), respectively. Both samples show good shielding performance, but the antioxidant properties of the two samples are obviously different. The average EMI SE of the Cu film decreased severely (from 44.55 dB to 27.69 dB) after heating due to the serious oxidation, while that of the Gr/Cu/Gr sample did not change much (from 44.17 dB to 41.12 dB). The maintenance rate of the average EMI SE of the Gr/Cu/Gr sample after heating at 120 °C for 3 h can reach 93.09%, far higher than that of the Cu film sample (62.15%), demonstrating the good oxidation resistance and EMI SE reliability of the Cu/Gr nanolayered composites. This means that the Cu/Gr nanolayered composites have high adaptability in the EMI shielding in certain harsh environment. This merit makes Cu/Gr nanolayered composites good candidate for EMI shielding in high temperature environment of next generation microelectronic device.

In order to further improve the THz shielding properties of the Cu/Gr nanolayered composites, we fabricated the 4Cu4Gr sample, which exhibits a very high average EMI SE of about 61 dB, as shown in Fig. 6(b). Moreover, with the development of flexible wearable electronic devices, shielding materials are required to have good flexibility and repeatedly bending property [11]. The graphene in the Cu/Gr nanolayered composites can effectively improve the resistance to fatigue during bending process due to the strong sp² bonding and high strength and modulus of graphene [16]. On the one hand, microcracks produced in one layer of Cu can be deflected by graphene at the Cu/graphene interface, which hinders the further propagation to the next layer of Cu. On the other hand, dislocations can be prevented by graphene from propagating across the Cu/Gr interface, which reduces the density of dislocations stacked at film/substrate interface [16]. To explore the flexibility and fatigue resistance properties, recurrent bending tests of the 4Cu4Gr and 4Cu samples on a tube with a diameter of 21.83 mm by a tweezer were performed as shown in Fig. 6(a) (details in Fig. S6). Both samples showed high EMI SE before bending experiments, as shown in Fig. 6(b). However, after 1500 bending cycles, the average EMI SE of the 4Cu sample decreased obviously (from 60.59 dB to 56.39 dB), while the average EMI SE of the 4Cu4Gr sample changed little (from 60.95 dB to 60.26 dB). This means that the maintenance rate of the SE value of the Cu/Gr nanolayered composite can reach 98.87% after 1500 bending cycles, obviously higher than that of multilayer Cu film (93.07%). Thus, the prepared Cu/Gr nanolayered composites behave much better stability and resistance to fatigue compared with multilayer Cu films without graphene. These characteristics endow the Cu/Gr nanolayered composites with promising prospects in EMI shielding, especially for flexible, portable and wearable devices.

In order to solve the problem that the existing THz shielding materials can not achieve excellent comprehensive performance at ultrathin thickness, Cu/Gr nanolayered composites with nanoscale thickness were constructed. And their EMI shielding properties, oxidation stability and flexibility were investigated comprehensively. From these results, it can be concluded that the Cu/Gr nanolayered composites exhibited excellent THz shielding performance with small thickness, good flexibility and high stability. Very high average EMI SE values of 44.17 dB, 54.56 dB, 58.34 dB and 60.95 dB were achieved at 0.1-1.0 THz with the thicknesses of only 40 nm, 80 nm, 120 nm and 160 nm, respectively. Among the previously reported materials, the Cu/Gr nanolayered composites exhibit an order of magnitude higher specific EMI SE compared with the best results in previous reports. Meanwhile, benefiting from the protection of graphene, the Cu/Gr nanolayered composite showed an excellent maintenance rate of 93.09% of the EMI SE after heating at 120 °C for 3 h in air, much higher than that of bare Cu thin film (62.15%), indicating the significant enhancement of oxidation resistance. Besides, there was no obvious decrease in the EMI SE of the 4Cu4Gr composite after 1500 bending cycles with a maintenance rate of 98.87%, much higher than that of 4Cu sample (93.07%), demonstrating the good flexibility and flexural fatigue resistance of Cu/Gr nanolayered composites. These exceptional properties endow the Cu/Gr nanolayered composites with great potential for high-performance EMI shielding application in next generation electric and communication devices.