Highly conductive copper films prepared by multilayer sintering of nanoparticles synthesized via arc discharge

The major challenges in producing highly electrically conductive copper films are the oxide content and the porosity of the sintered films. This study developed a multilayer sintering method to remove the copper oxides and reduce copper film porosity. We used a self-built arc discharge reactor to produce copper nanoparticles. Copper nanoparticles produced by arc discharge synthesis have many advantages, such as low cost and a high production rate. Conductive inks were prepared from copper nanoparticles to obtain thin copper films on glass substrates. As demonstrated by scanning electron microscopy analyses and electrical resistivity measurements, the copper film porosity and electrical resistivity cannot be significantly reduced by prolonged sintering time or increasing single film thickness. Instead, by applying the multilayer sintering method, where the coating and sintering process was repeated up to four times in this study, the porosity of copper films could be effectively reduced from 33.6% after one-layer sintering to 3.7% after four-layer sintering. Copper films with an electrical resistivity of 3.49 ± 0.35 μΩ·cm (two times of the bulk copper) have been achieved after four-layer sintering, while one-layer sintered copper films were measured to possess resistivity of 11.17 ± 2.17 μΩ·cm.


Introduction
In the electronics industry, electronic circuits and devices are conventionally manufactured using photolithography, vacuum deposition, and electroless plating. However, these traditional processes are multi-staged and require expensive facilities and environmentally-undesirable chemicals [1][2][3].
Printed electronics are attracting tremendous interest owing to low manufacturing costs, and operational simplicity [4]. Generally, printed electronics refer to a process in which printing technologies are applied to fabricate electronic circuits and devices [5]. One of the essential components in printable electronic devices is highly conductive patterns, such as conducting lines and films. Since nearly all the printed electronics are based on the principle of transferring inks to a substrate, various conductive inks, such as metal-based and carbon-based nanomaterials, have been extensively explored in the past decades.
Currently, conductive inks based on metal nanoparticles (NPs) or dissolved metal precursors have drawn increasing attention because of the high electrical conductivity of metals.
Highly conductive metals, such as silver (Ag, σ = 6.3 × 10 7 S m −1 ), copper (Cu, σ = 5.96 × 10 7 S m −1 ), gold (Au, σ = 4.42 × 10 7 S m −1 ), and aluminium (Al, σ = 3.78 × 10 7 S m −1 ), represent the primary system of choice. Ag-and Aubased conductive inks have been widely investigated and are already commercially available. Nevertheless, silver and gold are noble metals, and their application forlarge-scale production processes is limited by their great expense [5][6][7]. For this reason, copper has been considered as an alternative due to its low electrical resistivity and low cost. It has been demonstrated that the oxidation of Cu NPs is less rapid than that of Al NPs when exposed to ambient air [8,9]. In spite of this, the oxidation tendency of Cu NPs is still the major problem. Cu oxides would severely reduce the electrical conductivity of the printed patterns. Therefore, considerable research efforts are currently being devoted to minimizing the effect of oxidation. Generally, strategies to address the oxidation problem are based on exposing Cu nanoparticles (NPs) into an inert atmosphere [10,11], coating Cu NPs with a nonoxidizable shell [12][13][14][15][16][17], or eliminating the surface oxide layer during post-printing treatment [5,6,[18][19][20]. Our previous study showed the feasibility of preparing copper inks from nanoparticles synthesized by arc discharge. Compared with wet-chemical methods, Cu NPs generated by arc discharge have no stabilizing agents covering the surface of the Cu nanoparticle yet possess a high metallic purity. Nevertheless, these pure Cu NPs exhibit a high oxidation tendency when exposed to air. It has been shown that freshly collected Cu NPs exhibit an oxide layer with a thickness of about two nanometres. This oxide layer can be successfully removed by sintering Cu NPs in a reducing atmosphere, showing significant potential for the fabrication of printed electronics [6].
Apart from the oxidation problem, another factor governing the electrical resistivity is the thin-film porosity after sintering [21][22][23][24][25][26]. Compared with Cu patterns with a bulk structure, the porous structure of sintered copper patterns can significantly accelerate the oxidation process of printed patterns. To the best of our knowledge, only a few studies have addressed the problem of copper film porosity. Joo et al [27] have added Cu nanowires (NWs) in a Cu NP ink to obtain a densely packed network of Cu NW/NP structure, and thereby improving the electrical conductivity of the sintered Cu films. It was found that the porosity of flashlight-sintered Cu NW/ NP films was decreased to 4.07%, whereas the electrical resistivity of prepared Cu NW/NP films showed a low resistivity of 22.77 μΩ·cm. Rosen et al [28] used a combination of heat and pressure for sintering to obtain a dense copper layer. Copper patterns with equivalent specific resistivity as low as 5.3 ± 0.3 μΩ·cm were reported. The layer morphology and cross-sectional images demonstrated that a high temperature and pressure of 225°C and 1260 PSI (pound/inch 2 ) resulted in a denser copper layer. However, the relationship between the porosity evolution and the electrical resistivity of Cu films has not been discussed in detail. Chan et al [18] proposed a progressive three-step sintering process, including low-pressure drying, near-infrared sintering, and intense pulsed light (IPL) to ensure high nanoparticle compactness and low oxygen content. The porosity of the sintered Cu thin film was reduced to 6.3%, and the electrical resistivity of Cu films was decreased to 7.9 μΩ·cm. In our previous study, we applied formic acid as a reducing gas during thermal sintering to remove the copper oxide layer and form a connected metallic structure. Using this method, we could produce one-layer copper films with low resistivity of 5.4 ± 0.6 μΩ·cm [6]. Despite the continuous interconnection among nanoparticles, the porosity of sintered copper films after sintering has not yet been studied.
In this study, we focused on preparing a dense copper film to enhance the electrical conductivity of printed films. A multilayer-sintering method, in which the coating and thermal sintering processes were repeated up to four times, was developed. Using this process, we systematically studied the relationship between the porosity and electrical resistivity of the sintered copper films according to the obtained experimental results. First, the effect of one-layer film thickness on the electrical resistivity was investigated, where three different inks with 25 wt% Cu NPs were prepared to obtain copper films with various thicknesses. Furthermore, we investigated the porosity and electrical resistivity evolution of single-layer Cu films by prolonging the sintering time. By applying multilayer sintering, the film porosity and resistivity of different layers were experimentally studied and discussed. The copper layer morphology was analyzed using scanning electron microscopy, and the corresponding porosities were calculated by the image analysis software. The calculated porosity reduction is in good agreement with the decrease in the electrical resistivity of the film, indicating an effective strategy for fabricating highly conductive Cu patterns.

Fabrication of Cu NPs and Cu inks
The copper nanoparticles used in this study were produced using a laboratory-built arc discharge reactor. The reactor configuration and synthesis process have been described in our previous study [6,29]. Briefly, in the reactor chamber, bulk copper (copper shot, 0.8-2 mm, 99.5%, Alfa Aesar) was heated by a thermal plasma and vaporized into copper vapours. The generated copper atoms were immediately cooled by flowing inert gas (nitrogen with a purity of 99.9995%), which induced sequential processes, including nucleation, condensation, and coagulation, to form copper nanoparticles. In this work, the production parameters for Cu nanoparticles were the main carrier gas flow of 40 slm (standard liter per minute), cross gas flow of 3 slm, applied arc current of 30 A, and fixed electrode distance of 3 mm. A production rate of approximately 2 g h −1 of Cu NPs was achieved with these process parameters. The average primary particle size measured by a Brunauer-Emmett-Teller (BET) analyzer (Gemini VII 2390a, Micrometrics instruments corporation, GA, USA) was sub-100 nm. In this study, we used freshly synthesized Cu NPs (stored in a container within one week) for further ink production. Powder x-ray diffraction patterns were obtained using a Rigaku Smartlab High Resolution x-ray diffractometer equipped with a 9-kW rotating anode x-ray generator using a Cu-anode. Rietveld refinement using MAUD (materials analysis using diffraction) allowed us to confirm the crystal phases. The microstructure of the copper nanoparticles was observed via a scanning electron microscope (SEM, JSM 7500F, Jeol, Tokyo, Japan) with a secondary electron (SE) detector. Freshly produced Cu NPs were also studied using a transmission electron microscope (TEM, JEM-2200FS, Jeol, Tokyo, Japan).
The ink compositions are summarized in table 1. All prepared Cu inks have a metal loading of 25 wt%. When the spin-coating process parameter remained constant, the thickness of the spin-coated film depended on the solid material concentration and solvent properties, such as viscosity and evaporation rate. Therefore, we varied the ratio of glycerol to ethylene glycerol to obtain Cu films of different thicknesses. Ethanol, isobutanol, glycerol, and ethylene glycerol were all of the analytical grades and used without any purification. First, Cu NPs were added to a mixed solvent which is listed in detail in table 1. For the premixing process, the mixed Cu NPs and solvent were dispersed using a magnetic stirrer (MR Hei-Standard, Heidolph Instrument, Schwabach, Germany) for one hour at a stirring speed of 1000 revolutions per minute (rpm) and a hotplate temperature of 30°C. The premixed dispersions were then sonicated with an ultrasonic processor (UP100H, Hielscher Ultrasonics GmbH, Teltow, Germany) for one hour to break up the agglomerates, with an amplitude of 100 and a cycle of 0.9. Sonication was performed in an icewater bath to prevent overheating of the ink samples. All the prepared inks passed smoothly through the 0.7 μm syringe filter but did not flow through the 0.45 μm syringe filter, indicating that the agglomerates after sonication were between 450 nm and 700 nm. Moreover, the prepared inks remained stable for three days at room temperature. After three days, a re-sonication process of 15 min enabled the inks to be reused.

Sintering & characterization of copper films
Copper films were prepared using the spin coating technique (Spin 150, SPS Europe B.V., Putten, Netherlands). The prepared inks (10μl) were first added dropwise onto a glass substrate (10 mm×10 mm, Menzel Cover Glasses, Thermo Fischer Scientific, U.S.). For the spin coating process, all samples were rotationally accelerated to 7000 rpm and continuously spun for two minutes. Afterwards, a post-thermal process (sintering) was applied to remove the organic solvents and convert the copper particles into a continuous metallic state. The experimental setup for sintering has been demonstrated in our previous report, where nitrogen (a purity of at least 99.995%) was used as the carrier gas, and formic acid (analytical grade) was applied as the reducing agent [6]. In this work, all Cu films were sintered at 300°C under a diluted reducing atmosphere and the gas flows remained unchanged. The sintering time was varied to investigate its effect on the electrical resistivity of copper films. To reduce the film porosity, we first sintered one-layer copper films for one hour. Subsequently, this one-layer sintered sample was coated again using the spin coating technique and sintered for one hour to form a two-layer sintered sample. These processes were repeated to obtain three-or four-layered sintered film samples.
The electrical resistivity of the sintered Cu films was obtained from the sheet resistance and the Cu film thickness, which were determined using a four-point probe method (Keithley 4200 SCS, Keithley Instruments, Cleveland, USA) and a stylus profilometer (XP-2000, Ambios Technology, CA, U.S.), respectively. For each film sample, the film thickness and sheet resistance were measured at three different positions. The average thickness and sheet resistance were then calculated using these values. In addition, we utilized a secondary electron (SE) detector (SEM, JSM 7500F, Jeol, Tokyo, Japan) to observe the surface and cross-section microstructure of the sintered Cu films. Elemental analyses of the copper films were conducted using a scanning electron microscope (SEM, JSM 7500F, Jeol, Tokyo, Japan) equipped with an energy-dispersive x-ray (EDX) spectrometer (Quantax, Bruker Nano GmbH, Berlin, Germany) at 10 kV. Moreover, the porosity of the Cu thin films was calculated using an image analysis program (ImageJ) from the selected low-resolution SEM images (with two magnifications, 5000 x and 10 000 x). The SEM images were first converted to binary black-and-white images by setting a threshold. The area that is not occupied by the copper particles was identified as pores. Therefore, the surface porosity of the film was calculated by dividing the obtained pore area by the complete image area. In this study, at least five SEM images at each magnification were used to evaluate the average film porosity of an individual sample.

Results and discussion
3.1. Copper nanoparticles from arc discharge synthesis Figure 1(a) shows the microstructures of the Cu nanoparticles collected from typical synthesis. The resulting Cu NPs displayed a perfect spherical morphology unaffected by the primary particle size. As can be seen, this arc discharge synthesis leads to a broad particle size distribution, where primary particles range approximately from 10 to 150 nm. However, most of the Cu particles are smaller than 100 nm in primary size, and this result is in good agreement with the BET measurements, which indicate the average primary size of the Cu NPs is sub 100 nm. Based on the TEM image analysis results in figure 1(a), we confirmed that the resulting nanoparticles have a crystalline structure with a thin oxide shell on the particle surface. The thickness of the amorphous surface layer was approximately 2 nm, which might have been generated during the powder collection and sample preparation. The crystal structure of the Cu NPs was determined by analyzing the XRD pattern of the synthesized Cu NPs, as shown in figure 1(b). These diffraction peaks are in good agreement with the characteristic values of the (111), Generally, Cu films prepared by direct printing techniques possess low electrical conductivity owing to the presence of ink dispersants and pores among NPs. Therefore, post-printing treatment of Cu films is required to remove the ink solvents and to cause a continuous interconnection among the copper particles. In the present work, all prepared Cu films were thermally sintered at 300°C in a tube furnace under a reductive atmosphere. We investigated the electrical resistivity of the Cu films as a function of sintering time. For each sintering time, we prepared one-layer Cu films utilizing two different 25 wt% Cu inks (ink-1 and ink-2 in table 1), where the ratio of the dispersion solvent glycerol and ethylene glycerol was varied. As depicted in figure 2, the electrical resistivity decreased significantly from 34.4 μΩ·cm after 30 min to 10.8 μΩ·cm after 60 min for ink-1 and from 18.1 μΩ·cm after 30 min to 8.49 μΩ·cm after 60 min for ink-2, respectively. The resistivity of the sintered Cu films had its minimum value within one hour and remained almost constant, despite the prolonged sintering time from one hour to four hours for both inks. These results indicate that the copper nanoparticles are fully sintered, and the impurities, including pores and organic residues, cannot be further removed or minimized after sintering for one hour. Figure 3 shows optical images of the unsintered and sintered Cu films after various sintering times. The unsintered Cu film exhibited a dark brown color, which suggests that the Cu NPs were slightly oxidized during particle collection and ink preparation. The color of the sintered films changed from dark brown (unsintered), brown (30 min) to bronze-red (60 min). After one hour of sintering, the film color remained unchanged, which agrees that there was no further improvement in conductivity with increasing sintering time. In addition, no delamination was observed for the samples sintered for less than two hours. Nevertheless, various stages of delamination were observed depending on the film thickness, pre-existing cracks, and prolonged sintering time. It should be mentioned that cracks and delamination are also considered critical problems in intense pulsed light (IPL) sintering, and many researchers have dedicated themselves to solving these problems [18,21,30,31]. The reasons for delamination can  be summarized as follows. One is material shrinkage induced by a high temperature for a long time during the sintering process. When sintering Cu NPs on a rigid substrate, the material volume decreases, and material tensile stress arises through the thickness of the copper layer, which may cause crack formation. Another factor attributed to volume shrinkage is the thermal decomposition or evaporation of organic solvents. A long sintering process is necessary to provide sufficient time to evacuate the internal gas from the sintered film. To analyze the surface morphologies of the copper films after various sintering times, SEM micrographs were obtained, as shown in figure 4. Figure 4(a) shows an SEM image of the unsintered copper film after drying for 24 h in ambient air. Copper particles are present in agglomerates where spherically shaped primary particles are held together by van der Waals force. After 30 min of sintering, most of the Cu particles were connected to each other. At the same time, some NPs were not completely sintered, as some isolated copper islands were present, and the boundaries of the large pores were not smooth due to incomplete sintering. When prolonging the sintering time to one hour, we observed that a percolation path for the electricity was formed despite the presence of pores. Besides, the pore boundaries became smoother, revealing a complete sintering of the Cu particles. For films sintered after four hours, the Cu films also show well-sintered structures, where the number of pores >one μm slightly decreased and pores <one μm increased. According to Sherer and Garino [32], less porous film, as well as smaller pores, experience more stress from a rigid substrate. This explanation is in accordance with our observation of film delamination after sintering for four hours. To avoid film delamination, we sintered all samples for one hour in the following study.

Effect of one-layer copper film thickness on the morphology and resistivity
Previous studies applied copper inks with a metal loading of 40 wt%, and the average thickness of one-layer copper film was 1000 nm. The pores in the thick film were challenging to refill because they could not penetrate and reach the underlying layer. Thus, it is necessary to study the effect of singlelayer thickness on film resistivity. When the process parameters of spin coating remain constant, the thickness of the deposited Cu layer is affected by the properties of the coating inks, such as metal loading and ink viscosity. In this study, we produced three types of 25 wt% concentrated Cu inks. The ratio of the solvent glycerol and ethylene glycerol was varied to obtain inks with different viscosities, as summarized in table 1. Different printing methods have varying requirements for ink performance, such as printability, wettability, and adhesion to substrates. It should be noted that our current work focused mainly on producing copper films of various thicknesses. The adjustment of the ink performance for a specific need will be the focus of our future work. Generally, applying ink-1 enables us to obtain the thickest films of about 600 nm and ink-2 to have middle thick films of about 400 nm, whereas Cu films prepared from ink-3 possess the thinnest layer of approximately 200 nm. The electrical resistivity of the one-layer sintered copper films was investigated as a function of the film thickness, as depicted in figure 5(a), where all sintering temperatures and times were 300°C and one hour, respectively. Based on our raw data for single-layer sintered Cu films, we divided them into three categories based on their thicknesses: thin (100-300 nm), medium (300-500 nm), and thick (500-700 nm) films. The calculated average  electrical resistivity of thin, middle, and thick films was 15.3 μΩ·cm, 9.70 μΩ·cm, and 13.5 μΩ·cm, respectively. It was found that middle-thickness films had the lowest electrical resistivity. Figures 5(b) and 5(c) show the cross-sectional microstructures of the sintered Cu films with medium and thick thicknesses, respectively. It is noticeable that the Cu nanoparticles interconnect entirely in both conditions, while more pores are observed in a thicker layer. The pores inside the thicker films might have been caused by the unexhausted gas induced by the thermal decomposition of the ink constituents. Trapped pores are rarely found in thinner films (<500 nm) because decomposed organic additives can quickly reach the film surface. As a result, thicker films (500-700 nm) have a higher electrical resistivity than medium films. It is worth mentioning that the electrical resistivity of the sintered films exhibits a relatively large fluctuation. This fluctuation is because the agglomerate size and homogeneity of the inks needed to be optimized, resulting in uneven surface profiles and the formation of small cracks in the films.
To further investigate the reason for the varying resistivity, we conducted SEM measurements to analyze the surface microstructures of the films with various thicknesses, as illustrated in figure 6. Since silicon (Si) is the main element in the glass substrate, the presence of pores in the sintered Cu films can be verified via elemental mapping analysis (figures 6(d)-(f)). By comparing the elemental distribution of Si and Cu, we confirmed that the bright regions were sintered copper nanoparticles. Moreover, the Cu film with a thickness of 153 nm (figure 6(a)) displayed the most pores, most of which had an equivalent size of more than one μm. The highly porous structure of thin films is most likely the reason for their high electrical resistivity. As shown in figures 6(b) and (c), the copper layers have similar surface microstructures, where the size of most pores is smaller than 0.7 μm. In addition, it was observed that the ink layers were thick enough to cover the pre-existing cracks or voids; therefore, a continuous interconnection between Cu NPs was formed. In our previous study, Cu inks with 40 wt% metal loading were applied to produce single-layer films with a thickness of approximately 1000 nm. By comparing the x-ray diffraction (XRD) spectra of the Cu films before and after sintering, the diffraction peaks of copper oxides disappeared, illustrating that copper oxides were completely removed after sintering in a reducing gas flow [6]. Since all films prepared in this work  possess thickness smaller than 1000 nm, we assume that impurities resulting from copper oxidation can be neglected after reducing sintering for one hour at 300°C. Therefore, the different resistivities can be attributed to the varying porosity of the sintered film. As the film thickness increases beyond a critical thickness, many noticeable pores trapped in the copper film might cause a significant problem in disturbing the electron mobility and affecting the electrical conductivity. The critical thickness is mainly dependent on the properties of the ink, such as the metal loading, particle size, and homogeneity. In the case of our present experiments, the critical thickness should be approximately 400 nm, because the electrical resistivity of even thicker films tends to be higher.

Effect of multilayer sintering
XRD and EDX spectra demonstrated that impurities, including copper oxides and organic solvents in copper films, could be eliminated after sintering at 300°C in a reducing gas flow [6]. The film porosity is the primary factor that limits the minimum achievable electrical resistivity. As previously mentioned, the film porosity after sintering is the result of volume shrinkage caused by nanoparticle densification and evaporation of organic constituents. The film porosity cannot be significantly reduced by prolonging the sintering time or increasing the single-layer thickness. As shown in figure 6(a), the porosity of the thin Cu films after sintering is relatively large, and many pores possess diameters bigger than one μm. Since most of the copper agglomerates in the prepared inks were smaller than 0.7 μm, the large pores might be refilled by the second coating and subsequently sintering. To study the effect of porosity reduction, we repeated the coating and sintering process up to four times to obtain multilayer sintered copper films. Besides, to effectively fill the pores of the sintered copper thin films, the thickness of each layer should be as small as possible so that the Cu particles can penetrate deeply into the bottom of the film. In this regard, we applied ink-3 to prepare copper films with a thickness of approximately 200 nm. The resistivity evolution of the copper films after coating & sintering is illustrated in figure 7. The electrical resistivity decreased from 11.17 ± 2.17 μΩ·cm for onelayer sintered films to 3.49 ± 0.35 μΩ·cm for four-layer sintered films, which is two times the resistivity of bulk copper (1.72 μΩ·cm). In addition, the porosity reduction of the sintered films was consistent with the decreasing tendency of electrical resistivity. In the present study, we assumed the film surface porosity as the film porosity because metal films, after thermal sintering, have been demonstrated to possess a uniform structure and volume shrinkage throughout the film thickness [33]. The porosity of Cu films was reduced from 33.6% after single-layer sintering to 3.7% after four-layer sintering, revealing an approximately 30% reduction using the multilayer sintering method. Figure 8 shows the surface microstructures of the multilayer Cu films, along with the corresponding porosity calculation using image analysis software. It was confirmed that the films after four-layer coating and sintering had significantly lower porosity than the Cu films after one-layer coating and sintering. In addition, these microstructural observations demonstrate that the pores can be effectively refilled by re-coating and re-sintering when particles dispersed in ink are smaller than the pre-existing pores. We provided cross-sectional micrographs of Cu films after multilayer sintering to further clarify their densification. Figures 9(a) and (c) show the Cu films obtained after onelayer sintering, where ink-1 was applied to obtain films with a thick thickness (500-700 nm). Compared with single-layer sintered films, multilayer sintering leads to a uniform and dense microstructure. This denser packaging structure results in lower resistivity. In this study, most copper films with porosity <5% were obtained after four-time coating and sintering. As shown in figure 8(d), almost all pores after fourlayer sintering are smaller than 450 nm, indicating that these pores can only be filled with particles or agglomerates smaller than 450 nm. Nevertheless, the inks applied in this study have not yet been optimized, and the agglomerate size was between 450 and 700 nm. As the effectof oxygen content was eliminated during our previous study, this multiply sintering method can be applied to all kinds of inks and sintering methods. Further optimization of the ink quality, such as reducing the agglomerate sizes, could further reduce the film porosity. So far, the sintered copper thin films have achieved an average electrical conductivity of 2.9 × 10 7 S m −1 (50% of the bulk copper value) with an average porosity of 3.7%.

Conclusions
Copper nanoparticles are considered an alternative to silver and gold nanoparticles in printed electronics. Film porosity is one of the main factors affecting the electrical resistivity of printed copper films. Our previous work demonstrated that copper film impurities, such as copper oxides, can be eliminated after sintering in a reducing gas flow at 300°C. Based on this sintering method, we studied the relationship between the porosity and electrical resistivity of sintered copper films. Three different viscosities of 25 wt% Cu inks were fabricated and applied to prepare thin copper films via spin coating. These Cu inks allowed us to obtain copper films with various single-layer thicknesses ranging from 100 to 700 nm. It was found that the film porosity and electrical resistivity could only be reduced to a limited extent by varying the sintering time or increasing the one-layer film thickness. Because the generation of pores in the sintered copper films is inevitable, a multilayer sintering method, in which the step coating and sintering process were repeated up to four times in this work, was developed to reduce the film porosity and resistivity. We demonstrated that Cu films after four-layer coating and sintering had a low porosity of 3.7%, whereas single-layer sintered Cu films possessed an average porosity of 33.6%. Furthermore, the porosity reduction of the sintered copper films is in good agreement with the decrease in film electrical resistivity, indicating that this sintering method efficiently addresses the porosity issue. We could obtain copper films with an electrical resistivity of 3.49 ± 0.35 μΩ·cm, which is two times the resistivity of bulk copper, by applying the multilayer sintering method.