Multi-layered micro-patterns co-printed with Ag@CuO nano-ink for flexible devices

Silver (Ag) nano-ink is widely used to fabricate the micro-patterns of flexible electronic devices owing to its excellent conductivity and stability. However, the cost of micro-patterns prepared with silver nano-ink is high. Here, multi-layered, multi-inked (silver@copper oxide) micro-patterns were co-printed layer by layer using an in-house silver nano-ink and commercial copper oxide (CuO) nano-ink. The prepared micro-patterns were solidified by laser sintering. Among the co-printed micro-patterns, the micro-pattern with a top layer of copper (1CuO@2Ag) had the lowest resistance, which was 13.1 Ω cm–1. Additionally, 1CuO@2Ag had the smoothest topography and lowest porosity, which was attributed to effective sintering at the optimal laser power (3 W) for all co-printed micropatterns owing to the high absorptivity and low reflectivity of copper. Moreover, after 500, 1000, and 1500 cycles of fatigue testing, the resistance of 1CuO@2Ag increased by 1.5%, 8.4%, and 13.7%, respectively, indicating good reliability. The proposed method lays the foundation for further studies on Ag@CuO composites for micro-pattern preparation.


Introduction
Printed flexible electronic devices are light-weight and foldable, which are requisite features of wearable devices designed for intelligent sensing and monitoring [1][2][3][4].The core component is the micro-pattern, which is used as the conductive layer.With the rapid development of flexible electronics, the key process technologies of printed micro-patterns have become cost-effective [5][6][7] and efficient [8][9][10], with low-pollution [3] and lowwaste [11].Currently, most micro-patterns are printed using gold or silver, both of which have unique advantages, such as good conductivity and oxidation resistance [6,[12][13][14].Nevertheless, gold (Au) and silver are precious metals and thus costly.Therefore, manufacturing methods that can reduce this cost without sacrificing the ease and simplicity of the preparation processes are required.
Copper (Cu) is an economical material to fabricate micro-patterns because it is cheaper than gold and silver [15,16].In addition, copper has better light absorption properties than silver.However, Cu exhibits high resistance and is unstable owing to its susceptibility to oxidation [17][18][19].Placing copper particles on the top layer of the pattern can effectively improve the efficiency of laser sintering.To avoid oxidation and enhance stability, copper oxide ink is made from copper dissolved in a specific solvent, which is suitable for preparation and printing, and then copper oxide is reduced to copper after laser sintering, forming the micro-pattern.Yan et al [20] printed micro-patterns with copper oxide nano-ink prepared by thermal oxidation, and the results showed that the micro-pattern sintered by laser has a relatively low resistance and high adhesion, making it attractive for various applications, including functional layers and devices.Jones et al [21] proposed a direct laser write (DLM) process to fabricate micro-patterns by rapidly depositing CuO onto the substrate.The results showed that increasing the laser energy density lowers the resistance of the conductive pattern, thus increasing the conductivity.Xu et al [22] used the hydrothermal method to prepare CuO electrodes for supercapacitors.By applying energy density of 10.05 Wh kg −1 , they obtained CuO electrodes with a specific capacitance of 839.9 F g −1 , the highest value for the electrode, and a capacitance retention of 95.8% after 2000 charge-discharge cycles.Din et al [23] used copper oxide to prepare a flexible sensor with a high impedance for humidity and pressure detection.Dawoon et al [24] applied post-heat treatment to porous copper films to prepare gas sensors for the detection of H 2 S. The results showed that increasing the porosity increases not only the response sensitivity of the sensor but also the conductivity of the copper oxide film.Hu et al [25] used pulsed laser deposition (PLD) at different oxygen pressures to prepare CuO films with different conductivities.According to the above research results, copper oxide is a feasible material for manufacturing micro-patterns.Nevertheless, the resistance of micro-patterns printed with copper oxide is not ideal for conductive layers of flexible devices.
In this work, to improve copper oxide nano-ink for the fabrication of high-performance micro-patterns applied to flexible devices, a method was developed to co-print multi-layered, multi-inked (silver@copper oxide) micro-patterns on polyethylene terephthalate (PET) films, which were subsequently laser sintered.Additionally, we analyzed the performances of direct-written micro-patterns with different co-printed structures.The resistance and conductivity of the prepared micro-patterns were measured using a digital source meter, and the surface topography of the prepared micro-patterns was determined using scanning electron microscopy (SEM) and optical microscopy.Furthermore, the reliability of the prepared micro-patterns was evaluated using fatigue tests.

Preparation of silver nano-ink
The liquid-phase reduction approach was used to prepare the silver nano-ink, as reported in our previous work [26].Especially, the ink prepared is a silver nanoparticle ink with multi-particle sizes distribution with bimodality, which is improved based on the single particle size nano-ink.Meanwhile, the size distribution of nano-particle ranges from 20 to 50 nm, and the solid content reaches 16.5 wt%, and the viscosity is 2 Cps, meeting the usage requirements for printing.In brief, the best advantage of the ink prepared is to make the micro-patterns denser, reducing the resistivity of the micro-patterns printed.Significantly, the silver nano-ink used in this work is not different from the ink reported previously.

Preparation of multi-layered, multi-inked micro-patterns
The printed pattern was prepared on a film substrate using the same printing system detailed in previous works [12,27], as shown in figure 1.As shown in figure 1(II, III), the silver nano-ink and copper oxide nano-ink were ejected from a nozzle with an inner diameter of 0.3 mm to generate the micro-pattern on the PET film layer by layer.Printing using this equipment was controlled by the C++ program.In addition, the parameters of printing were ink flow rate of 0.5 ml h −1 , applied voltage of 2600 V, and moving speed of 45 mm/s.After printing, the conductive pattern was laser sintered to obtain the programmed micro-pattern, as shown in figure 1(IV).
To optimize the conductivity of the micro-pattern, three micro-patterns were printed.They were (1) two layers of silver on a layer of copper (2Ag@1CuO), (2) a layer of silver on a layer of copper on a layer of silver (1Ag@1CuO@1Ag), and (3) a layer of copper on two layers of silver (1CuO@2Ag).Additionally, two control micro-patterns were printed for comparison.They were (1) three layers of silver (3Ag) and (2) three layers of copper (3CuO).Each micro-pattern had the same size (length: 1.5 mm, width: 200 μm, height: 50 μm).The resistance of the micro-patterns was measured using a digital source meter (Keithley 2450).Additionally, to reduce the measurement error, three samples of each micro-pattern were prepared.The size of the sample was measured using a step meter (DektakXT, Bruker, Karlsruhe, Germany).

Characterization
The surface topography of the micro-patterns was determined using SEM (S-4800, Hitachi, Tokyo, Japan) and optical microscopy (DMI3000M/DFC450, Leica, Wentzler, Germany).The crystalline structure of the film samples was examined using x-ray diffraction (XRD; D8Quest, Bruker, Saarbrucken, Germany).The resistance of the micro-patterns was measured using a digital source meter (Keithley 2450).

Effect of laser power on the conductivity of micro-patterns
The cross-sections of the micro-patterns are shown in figure 1(a).The micro-patterns were printed and needed to be solidified by laser sintering, which was performed with the following parameters: The height between the laser head and the PET film was set to 5 cm, the laser pulse width was 150 μm, the laser frequency was set to 7 Hz, and the sintering moving speed was 10 mm/s.To determine the effect of the laser power on the conductivity of the micro-pattern, the different micro-patterns were sintered at different laser powers, specifically 1, 2, 3, and 4 W. The resistances of five samples of the same sintered micro-pattern were measured, and the values were averaged, as shown in figure 2(b).
The results (figure 2(b)) showed that the resistances of 2Ag@1CuO, 1Ag@1CuO@1Ag, and 1CuO@2Ag first decreased and then increased as the laser power increased from 1 to 4 W, reaching the lowest value at 3 W, which were 18.7, 16.2, and 13.1 Ω cm -1 , respectively.This trend indicated that the silver and copper particles were not completely sintered when the laser power was less than 3 W.In contrast, the micro-patterns were ablated (figure 2(h)) when the laser power was greater than 3 W, leading to an increase in the resistance except in the case of 3Ag.Additionally, the resistance changes of 3Ag and 3CuO were significantly different from those of the micro-patterns co-printed with Ag@CuO nano-ink.This discrepancy can be explained as follows: Each micropattern printed with Ag and CuO has the same surface area but different heat absorptivity [28][29][30] under the same light source.Meanwhile, the heat absorptivities of the Ag and CuO are 9% [31] and 25% [32], respectively.The thermal conductivity of the Ag is higher than CuO, however, CuO film could absorb more energy than Ag film at the same laser energy.Specifically, light absorption per unit area and thus the heat absorptivity of Ag is lower than that of CuO, as shown in figures 3(a)-(b).As a result, solidification of 3Ag required a greater amount of energy and thus a higher laser power than solidification of 3CuO.For this reason, the resistances of 2Ag@1CuO, 1Ag@1CuO@1Ag, and 1CuO@2Ag were between the resistances of 3CuO and 3Ag.The surface topography also affected the resistivity of the prepared film.So, the surface topographies of 2Ag@1CuO, 1Ag@1CuO@1Ag, 1CuO@2Ag, 3Ag and 3CuO sintered at 4 W, were observed by the optical microscopy at a magnification of 1000, shown in figures 3(c)-(g).As shown in figure 2(e), the surface topography of 1CuO@2Ag after laser sintering was the smoothest, while the surface topography of 2Ag@1CuO and 1Ag@1CuO@1Ag were smoother than that of 3Ag and 3CuO.To further study the ablation degree of 3CuO sintered at 4 W, the surface topography, observed by the optical microscopy at a magnification of 1500, was presented in figure 3(h).Unfortunately, the film of 3CuO had been ablated severely, and unusable.

Effect of copper layer on performance of micro-pattern
To study the effect of the copper layer on the sintering and conductivity of the micro-patterns, 2Ag@1CuO, 1Ag@1CuO@1Ag, and 1CuO@2Ag were sintered completely.To demonstrate effective utilization of copper oxide and silver during sintering, SEM images of the micro-patterns and XRD patterns of 1CuO@2Ag were acquired, as shown in figure 3.
As shown in figures 3(c)-(e), the surface topography of 1CuO@2Ag was the smoothest.Owing to the considerably higher heat absorptivity of CuO than that of Ag, 1CuO@2Ag absorbed a greater amount of laser energy under the same laser power than the other micro-patterns.For this reason, most of the laser energy absorbed by the copper oxide particles on the upper layer was transmitted to the silver layer, which increased the rate of sintering, and 1CuO@2Ag was completely solidified.In addition, according to previous studies [33][34][35][36], copper oxide can absorb up to 90% of the laser energy and reflects only 10% of the laser energy.However, if the silver layer was printed as the upper layer, most of the laser energy was reflected, as shown in figures 3(a) and (b).
We also investigated the effect of porosity on the performance of the micro-pattern.The porosities of 2Ag@1CuO and 1Ag@1CuO@1Ag reached 75.8% and 56.7%, respectively.In contrast, the porosity of 1CuO@2Ag was only 8%.Notably, 1CuO@2Ag had a lower resistance than the other micro-patterns.The porosity was determined by calculating the number of pores per unit area (200 × 200 μm) using optical Relationship between the resistance of the micro-pattern and laser power.Optical microscopy images of (c) 2Ag@1CuO, (d) 1Ag@1CuO@1Ag, (e) 1CuO@2Ag, (f) 3Ag, and (g) 3CuO after laser sintering at 4 W, observed at a magnification of 1000.(h) Optical microscopy image showing ablation of the 3CuO sintered at 4 W, observed at a magnification of 1500.microscopy images.Although the resistance of 1CuO@2Ag was higher than that of bulk silver, using the multiink could effectively reduce the cost of manufacturing micro-patterns for flexible electronics.
The XRD pattern (figure 3(f)) indicated that sintering 1CuO@2Ag completely at 3W produced the best crystalline film.XRD characterization of the flexible PET substrate was also carried out for comparison.As shown in figure 3(f), three diffraction peaks at 13°, 40°, 49°corresponded to (110), (002), and (111) of the CuO particles crystal plane (JCPDS # 48-1548), respectively.Additionally, three diffraction peaks at 27°, 44°, 70°c orresponded to (200), (311), and (220) of the Ag particles crystal plane (JCPDS # 04-0783), respectively.As the 2θ angle increased, the intensity of the CuO and Ag peaks decreased because the thickness of the film close to the edge decreased.The Talysurf Profiler (Dektak-XT-10th, Bruker, Massachusetts, USA) was applied to study the height of the 1CuO@2Ag film, shown in figure 3(g).It could be seen that the average height of the 1CuO@2Ag film was 63 μm, and its height fluctuation changes was small.

Reliability analysis
To evaluate the reliability of 1CuO@2Ag after sintering, 1500 cycles of the bending fatigue test was conducted with the micro-pattern fixed on the fatigue testing machine, as shown in figure 4.
After 500, 1000 and 1500 cycles, the resistance of the micro-pattern was measured three times, and the average resistance R was obtained to reduce the measurement error, shown in figure 5.
As shown in figure 2(b), the resistance R 0 was 13.1 Ω cm -1 before the bending fatigue test.The change in resistance S was used as an indicator of the reliability of the micro-pattern.The change in resistance S was calculated as follows: As shown in figure 5(a), the resistance of the micro-pattern after the fatigue test and the change in resistance S both showed an upward trend with increasing cycles of the fatigue test.Nevertheless, the resistance of the micropattern R after 1500 cycles of the fatigue test was only 14.9 Ω cm -1 , and the change in resistance S was 13.7%.Moreover, the LED lamp could be lit.Furthermore, the adhesion strength of the micro-pattern to PET substrate is very important to applications.So, the 500 cycles of peel-off experiment were conducted by adhesive cellophane tape, and the resistivity of 1CuO@2Ag was measured three times after every 100 cycles, and then the average resistivity was obtained, shown in figure 5(c).From the figure 5(c), it could be seen that the resistivity of the micro-pattern was increased by 8.3%, compared with 13.1 Ω cm -1 .These results indicated that the reliability of 1CuO@2Ag was good.
To demonstrate the availability of the micro-pattern of 1CuO@2Ag printed, the comprehensive comparison of micro-pattern prepared in this work was compared with other inks printed by different printing methods and sintered by different sintering methods, shown in table 1.It could be seen that the resistivities of micro-patterns printed by Ag and Cu@Ag inks [37,39,40] respectively, were higher than that of this work.However, the resistivity of micro-patterns printed by CuO@Ag ink [41] is lower, compared with this work.It is the reason that  the micro-pattern was sintered by hot-pressing, leading to improving the density of micro-patterns and reducing the resistivity.Worth mentioning, the resistivity of micro-pattern in this work was close to the result of Ag film [38].Therefore, 1CuO@2Ag can meet the requirements stipulated by the flexible electronic industry for flexible sensor devices.

Conclusions
Multi-layered, multi-inked micro-patterns were printed layer by layer on PET films using a self-synthesized silver nano-ink and commercial copper oxide ink.Compared with the micro-pattern printed using silver only, the micro-pattern printed using the proposed co-printing method required a reduced laser power to solidify, which improved the efficiency of manufacturing micro-patterns.In particular, owing to the high laser-energy absorptivity of copper oxide, which was printed as the upper layer of 1CuO@2Ag, the surface topography, resistance, and porosity of 1CuO@2Ag were better than those of 2Ag@1CuO and 1Ag@1CuO@1Ag.Although the resistance of 1CuO@2Ag was higher that of 3Ag after complete sintering, it was sufficiently low for flexible devices.Furthermore, after 500, 1000, and 1500 cycles of the fatigue test, the resistance of 1CuO@2Ag increased by 1.5%, 8.4%, and 13.7%, respectively.Therefore, co-printing multi-layered, multi-inked micro-patterns using silver and copper oxide is a potential method for fabricating low-cost flexible devices.

Figure 1 .
Figure 1.Schematic of the preparation process of the micro-patterns.

Figure 2 .
Figure 2. (a) Cross sectional images of different micro-patterns.(b)Relationship between the resistance of the micro-pattern and laser power.Optical microscopy images of (c) 2Ag@1CuO, (d) 1Ag@1CuO@1Ag, (e) 1CuO@2Ag, (f) 3Ag, and (g) 3CuO after laser sintering at 4 W, observed at a magnification of 1000.(h) Optical microscopy image showing ablation of the 3CuO sintered at 4 W, observed at a magnification of 1500.

Figure 4 .
Figure 4. (a) Principle of the bending cycle fatigue test.(b) Physical image of the fatigue testing machine.

Figure 5 .
Figure 5. (a) Resistance of 1CuO@2Ag after the fatigue test and change in resistance S. (b) Conductivity of 1CuO@2Ag after 1500 cycles.(c) Resistance of the micro-pattern after 500 cycles of peel-off test.