Research Article www.acsami.org
Intensive Plasmonic Flash Light Sintering of Copper Nanoinks Using a Band-Pass Light Filter for Highly Electrically Conductive Electrodes in Printed Electronics Yeon-Taek Hwang,† Wan-Ho Chung,† Yong-Rae Jang,† and Hak-Sung Kim*,†,‡ †
Department of Mechanical Convergence Engineering, Hanyang University, 17 Haendang-Dong, Seongdong-Gu, Seoul 133-791 South Korea ‡ Institute of Nano Science and Technology, Hanyang University, Seoul, 133-791 South Korea ABSTRACT: In this work, an intensive plasmonic flash light sintering technique was developed by using a band-pass light filter matching the plasmonic wavelength of the copper nanoparticles. The sintering characteristics, such as resistivity and microstructure, of the copper nanoink films were studied as a function of the range of the wavelength employed in the flash white light sintering. The flash white light irradiation conditions (e.g., wavelength range, irradiation energy, pulse number, on-time, and off-time) were optimized to obtain a high conductivity of the copper nanoink films without causing damage to the polyimide substrate. The wavelength range corresponding to the plasmonic wavelength of the copper nanoparticles could efficiently sinter the copper nanoink and enhance its conductivity. Ultimately, the sintered copper nanoink films under optimal light sintering conditions showed the lowest resistivity (6.97 μΩ·cm), which was only 4.1 times higher than that of bulk copper films (1.68 μΩ·cm) KEYWORDS: copper nanoparticle, flash light sintering, intensive plasmonic, optical filter, band-pass filter, wavelength range, printed electronics
■
INTRODUCTION Printed electronics technology has recently been receiving increased attention because it can minimize facility needs and energy consumption, as well as reduce the emission of environmentally toxic materials during patterning. Therefore, printed electronics technology is an attractive alternative to conventional photolithography and can be applied in many areas including flexible displays, flexible solar cells, wearable electronics, organic thin-film transistors (OTFT), and organic light emitting diodes (OLED).1−8 Within the realm of printed electronics technology, metal nanoparticle inks have been studied for wide usage. Previously, in the area of noble metals, attention has been focused on silver nanoparticles which have been commonly used as a conductive material because of their superior oxidation stability and conductivity.9−12 However, the cost of silver nanoparticles is quite high. As a substitute for silver nanoparticles, copper nanoparticles have received considerable attention because of their low cost and high electrical conductivity.13−15 Unfortunately, copper nanoparticles are plagued by several problems; they can be easily oxidized at room temperature and cannot be sintered by thermal sintering under ambient conditions.16,17 A number of sintering methods (e.g., laser sintering, microwave sintering and plasma sintering)18−21 were studied to solve these problems. However, these approaches have limitations in mass production because of their low throughput, high complexity, and considerable environmental obstacles, often requiring high © XXXX American Chemical Society
temperature or vacuum conditions with reducing or noble gases (hydrogen or nitrogen). The flash white light sintering method has been developed to address these problems.22−24 This method can easily reduce the copper oxide shell of poly(N-vinylprrolidone) (PVP)-coated copper nanoparticles because the oxide shell of copper nanoparticles were removed by the intermediate alcohol from decomposed PVP25 and sinter a copper nanoink film of a large area without damaging the polymer substrate, all in a few milliseconds at room temperature under ambient conditions. However, in previous studies, the resistivity of the sintered copper film is still high compared to the bulk copper.26,27 Also, it is not clear how the wavelength in the white flash light affects the sintering characteristics of the copper nanoparticles. In this study, the effect of wavelength of the white flash light on the sintering characteristics of the copper nanoink was investigated using various light filters (e.g., high pass, low pass, and band-pass filters). Specific wavelength range was passed or blocked by various light filters. In addition, the flash white light irradiation conditions (e.g., pulse duration, irradiation energy, pulse numbers) were optimized to improve the geometrical and electrical characteristics of the copper nanoink films. Received: December 22, 2015 Accepted: March 15, 2016
A
DOI: 10.1021/acsami.5b12516 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
Figure 1. (a) Schematics of the optical filter applied flash light sintering system, and wavelength spectrum of the flash light (b) without filter and with (c) 500 nm high-pass filter, (d) 600 nm low-pass filter, and (e) 500−600 nm band-pass filter.
Figure 2. (a) Resistivity of the flash light sintered copper nanoink film without filter (on-time, 5 ms; pulse number, 1) and SEM images of the flash light sintered copper nanoink film with respect to irradiation energy of (b) 3.5, (c) 4.0, (d) 4.5, and (e) 5.0 J/cm2.
prepared copper nanoink film according to the irradiation energy without a filter. Figure 2a shows the resistivity of the sintered copper nanoink film for irradiation energy ranging from 3.5 J/cm2 to 5.0 J/cm2 (pulse number, 1; on-time, 5 ms; Table 1). The sintered copper nanoink films under an irradiation condition of 4.0 J/cm2 had the lowest resistivity because their necking was well formed (Figure 2c). Meanwhile, for an irradiation energy over 4.0 J/cm2, the resistivity of sintered copper nanoink films increased because of delamination and cracking that occurred on the surface of the copper nanoink films by the excessive flash light irradiation energy (Figures 2d-e). Therefore, the irradiation energy to sinter the copper nanoink films was chosen to be 4.0 J/cm2. The wavelength of the flash light was selectively chosen and used with a high-pass, low-pass, and band-pass filters (Table 1).
Moreover, to analyze the sintering mechanism according to the wavelength range, in situ monitoring of temperature was performed using a noninverting amplifier circuit including an op-amp and a thermocouple to monitor the sintering mechanism of the copper nanoparticles.
■
RESULTS AND DISCUSSION As shown in Figure 1, the flash white light emitted from the xenon lamp could sinter the copper nanoink films. The wavelength of the flash white light (350−950 nm) was controlled by various light filters (high-pass filter, low-pass filter, and band-pass filter) and its wavelength distribution was analyzed by a spectrometer (Figure 1b−e). To find the optimal sintering conditions of the copper nanoink films, the flash white light was irradiated on the B
DOI: 10.1021/acsami.5b12516 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces Table 1. Selective Wavelength Flash Light Sintering Conditions pulse no.
on-time (ms)
irradiation energy (J/cm2)
wavelength range (nm)
no filter
1
5
350−950
high-pass filter
1
5
3.5 4.0 4.5 5.0 4.0
low-pass filter
1
5
4.0
band-pass filter
1
5
4.0
430−950 500−950 600−950 700−950 350−700 350−600 350−500 500−600
The various high-pass filters (430, 500, 600, and 700 nm) which can pass through the high wavelength range and block the low wavelength range were applied to the bottom of the xenon lamp (Figure 1a), and its wavelength distribution was measured by a spectrometer (Figure 3a−d). The flash light irradiation energy was measured by a power meter (Nova II power meter, OPHIR Optronics Solutions, Ltd.) and precisely recalibrated in all cases. Note that the overall light energy of the flash light was changed by the wavelength filter, even though the irradiation energy from the xenon lamp itself was the same. As shown in Figure 3e, the resistivity of copper nanoink films made with the 430 nm highpass filter and 500 nm high-pass filter was lower than that of the sintered copper-ink films without a filter. Also, it was determined that by using a 500 nm high-pass filter, the neck-like junctions among the copper nanoparticles grew uniformly with small pores (Figure 3h). Meanwhile, in the case of the high-pass filter of 600 and 700 nm, the resistivity increased again. As shown in the SEM images of Figure 3i,j, at the applied flash light with a 600 or 700 nm high-pass filter on the copper nanoink films, the necking among the copper nanoparticles was not uniformly formed with high porosity. The temperature change of the copper nanoink film during the flash light process was monitored in each of the high-pass wavelength filter cases (Figure 4). It was noteworthy that in the same irradiation conditions (e.g., irradiation energy, on-time,
Figure 4. In situ monitoring of the temperature changes of copper nanoink films (a) without filter and with high-pass filters of (b) 430, (c) 500, (d) 600, and (e) 700 nm (irradiation energy, 4.0 J/cm2; on-time, 5 ms; pulse number, 1).
pulse number), the maximum temperature of the copper nanoink films of the 500 nm high-pass filter case was the highest among all other cases (Figure 4c). Meanwhile, in the 600 and 700 nm high-pass filter cases, the maximum temperature of the copper nanoink films was significantly lower than that of the other two high-pass filter cases (430 and 500 nm high-pass filters). The maximum temperature was even lower than the case of no filter (Figure 4d,e). This phenomenon might arise from the light absorption wavelength range of the copper nanoparticles. Figure 5 shows the UV−vis results of the copper nanoparticles, where a light absorption band of copper nanoparticles existed around 590 nm. Note that the temperature of the copper nanoink films during the flash light sintering process without a wavelength of 590 nm was much lower than those with a wavelength of 590 nm (Figure 4). Therefore, it might be concluded that the surface plasmon phenomena of the copper nanoparticles contributes mainly to its flash light sintering phenomena. To study this further, various low-pass filters (e.g., 700 nm, 600 and 500 nm), which block the high wavelength range, were
Figure 3. Wavelength spectrum for high-pass filter of (a) 430, (b) 500, (c) 600, and (d) 700 nm. (e) Resistivity of the flash light sintered copper nanoink film without filter and with high-pass filter (irradiation energy, 4.0 J/cm2; on-time, 5 ms; pulse number, 1) and SEM images of flash light sintered copper nanoink film (f) without filter and with high-pass filters of (g) 430, (h) 500, (i) 600, and (j) 700 nm. C
DOI: 10.1021/acsami.5b12516 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
As shown in the SEM images of Figures 6e−g, the copper nanoparticles were more firmly connected with the 700 and 600 nm low-pass filters than the case where a filter was not used. This figure also reveals that the flash light which included the copper nanopaticle’s plasmonic wavelength of 590 nm (700 and 600 nm low-pass filtered light) could more effectively sinter the copper nanoink (Figures 6a,b). Meanwhile, light passing through the 500 nm low-pass filter (Figure 6c), could not efficiently sinter the copper nanoink film, and its resistivity became abruptly high, as shown in the SEM image of Figure 6h. The results obtained while monitoring temperature during the low pass filtered light are shown in Figure 7. It was found that the temperature of the copper nanoink film during the flash light sintering process was lower than 150 °C in the case of the 500 nm low-pass filtered light sintering (Figure 7d). Meanwhile, when the 600 nm lowpass filtered light was irradiated, the maximum temperature was increased up to 200 °C, which was higher than any of the other low-pass filter cases (Figure 7c). From these results, it was found that the flash light including a 600 nm wavelength, could heat the temperature of the copper nanoparticles to a high temperature more efficiently due to the plasmonic effect of the copper nanoparticles (Figure 5). It is worth noting that the results of the 600 nm low-pass filter case were similar to that of the 500 nm high-pass filter case, as the high influence of plasmonic wavelength of copper nanoparticles (500−600 nm) were irradiated in both cases (Figure 3).
Figure 5. UV−vis spectra of copper nanoparticles.
used during the flash light irradiation (Figures 6a−c). As shown in Figure 6d, there was a change of resistivity in the copper nanoink film according to the wavelength range of the low-pass filter. Using a low-pass filter of 700 nm, the resistivity of the sintered copper nanoink film was lower than that of the no filter case. Moreover, for the case with the 600 nm low-pass filter, the sintered copper nanoink film had the lowest resistivity observed.
Figure 6. Wavelength spectrum for low-pass filter of (a) 700, (b) 600, and (c) 500 nm. (d) Resistivity of the flash light sintered copper nanoink film without filter and with low pass filter (irradiation energy, 4.0 J/cm2; on-time, 5 ms; pulse number, 1), and SEM images of flash light sintered copper nanoink film (e) without filter and with low-pass filters of (f) 700, (g) 600, and (h) 500 nm. D
DOI: 10.1021/acsami.5b12516 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
the flash white light using a band-pass filter (from 500 to 600 nm). Note again that the total flash light energy was constant, thus the influence in the 500−600 nm range band-pass filtered light was higher than those of the other cases because the overall light energy of the flash light was concentrated at the wavelength range between 500 and 600 nm. As shown in Figure 8d, the resistivity of the sintered copper nanoink films was the lowest when the band-pass filtered flash white light (500−600 nm) was used. Also, as shown in the SEM image in Figure 8, the band-pass filtered copper nanoink film (Figure 8h) had a denser necking structure among the copper nanoparticles than those sintered by a low-pass or high pass filtered flash light (Figures 8f,g). It was also found that the maximum temperature of the copper nanoink films during flash light irradiation with the band-pass filter (243 °C) was higher than that with the high-pass (216 °C) and low-pass (212 °C) filters, as shown in Figure 9d. This is because the flash light with a band-pass filter could intensively irradiate the light in the plasmonic wavelength range of the copper nanoparticles, namely the intensive plasmonic flash light sintering. The XRD patterns of the copper nanoink films, both unsintered and sintered, as well as without a filter and with various filters, are shown in Figure 10. The XRD patterns of the unsintered copper nanoink film showed a copper(I) oxide peak at 36.4°. When 4.0 J/cm2 of flash light without a filter was irradiated, the copper
Figure 7. In situ monitoring of the temperature changes of the copper nanoink films (a) without filter, with low-pass filter of (b) 700 nm, (c) 600 nm, and (d) 500 nm. (irradiation energy: 4.0 J/cm2, on-time: 5 ms, pulse number: 1).
The band-pass filter ranging from 500 to 600 nm was used and compared to the 600 nm low-pass and 500 nm high-pass filter mentioned above. Figure 8c shows the wavelength spectrum of
Figure 8. Wavelength spectrum for (a) 500 nm high-pass filter, (b) 600 nm low-pass filter, and (c) 500−600 nm band-pass filter. (d) Resistivity of the flash light sintered copper nanoink film without filter and with high-pass filter, low-pass filter and band-pass filter (irradiation energy, 4.0 J/cm2; on-time, 5 ms; pulse number, 1), and SEM images of flash light sintered copper nanoink film (e) without filter and with (f) 500 nm high-pass filter, (g) 600 nm low-pass filter, and (h) 500−600 nm band-pass filter. E
DOI: 10.1021/acsami.5b12516 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
films at the same flash light energy conditions was enhanced. Therefore, as shown in Figure 11, the sintering characteristic of the copper nanoparticles was improved by the selective wavelength flash light sintering process using a band-pass filter from 500 to 600 nm. The irradiation conditions, such as pulse number, the on-time and the off-time, were further controlled and optimized. The intensive plasmonic flash light process filtered by a bandpass filter (500−600 nm) was divided into two steps: a preheating step and a main sintering step (Table 2). The preheating step is known to eliminate the organic binder and the main sintering step is used to complete the necking connections among the copper nanoparticles.31 In this work, the pulse number, the on-time and the off-time of the preheating step were fixed at 30, 1, and 30 ms, respectively, while the preheating light energy was varied from 4 J/cm2 to 10 J/cm2. Figure 12 shows the resistivity and SEM image of two-step flash light sintered copper nanoink films with respect to the preheating energy. When a preheating irradiation condition of 8 J/cm2 was employed, the sintered copper nanoparticles were heavily agglomerated and had the lowest resistivity (Figure 12e). Further, the sintered copper nanoparticles showed denser necking connections and obviously lower porosity than any of the other cases. Finally, it was found that intensive plasmonic flash light sintering with a two-step process and band-pass filter could achieve a highly conductive copper nanoink film with a low resistivity (6.97 μΩ·cm). This value was 4.1 times higher than that of the bulk copper films (1.68 μΩ·cm). Figure 13 shows the printed copper nanoink patterns before and after flash light sintering. With flash light irradiation without filter, the copper pattern was partially sintered due to the low temperature of the copper nanoink film (175 °C) during the sintering process (Figure 13b). However, when the intensive plasmonic flash light has been irradiated, the copper pattern was fully sintered because the plasmonic wavelength, which induces photothermal heating, could sinter the copper pattern at a high temperature (243 °C) (Figure 13c).
Figure 9. In situ monitoring of the temperature changes of copper nanoink films (a) without filter and with (b) 500 nm high-pass filter, (c) 600 nm low-pass filter, and (d) 500−600 nm band-pass filter (irradiation energy, 4.0 J/cm2; on-time, 5 ms; pulse number, 1).
■
Figure 10. X-ray diffraction patterns of copper nanoink films for unsintered and flash light sintered with various filter cases (irradiation energy, 4.0 J/cm2; pulse duration, 5 ms; pulse number, 1).
CONCLUSIONS In this work, a selective wavelength during flash light sintering of the copper nanoink was investigated using various filters (highpass, low-pass, and band-pass filters). To improve the sintering characteristics of the copper nanoparticles, we optimized the wavelength range of the flash white light under constant irradiation conditions (4 J/cm2 energy, 1 pulse, 5 ms on-time). Copper nanoink films were found to be efficiently sintered by the band-pass filter (from 500 to 600 nm) applied flash light due to the concentrated irradiation of the wavelength range, which included the absorption wavelength of the copper nanoparticles (590 nm). The copper nanoink films sintered by the intensive plasmonic flash light with a two-step process (8 J/cm 2 preheating, 4 J/cm2 main sintering energy) had a low resistivity of 6.97 μΩ·cm, which is 4.1 times higher than that of bulk Copper films (1.68 μΩ·cm). It is expected that the results obtained in this work can be widely used to improve the sintering characteristic of metallic nanoparticles that have a specific absorption wavelength range.
oxide peak decreased and the intensity of the copper phase peaks (43.2, 50.4, and 74.1°) increased, which is similar to the results of our previous study.25 Nevertheless, the copper oxide peak still remained. The oxide shells of copper nanoparticles were removed by the intermediate alcohol from decomposed PVP during a few milliseconds. Also, for decomposition of the PVP binder to reduce copper oxide, a temperature of over 200 °C was required.28 However, for the flash light irradiated without a filter, the sintering temperature of the copper nanoink films was about 175 °C. Therefore, the copper oxide surrounding the copper nanoparticles was not completely reduced. Meanwhile, after flash light irradiation with the filters (e.g., 500 nm high-pass filter, 600 nm low-pass filter and 500−600 nm band-pass filter), the copper oxide peak was completely removed by full decomposition of the PVP binder because the sintering temperature of the copper nanoparticles was increased up to 200 °C. Figure 11 shows a schematic of the intensive plasmonic wavelength flash light sintering. From this study, it was found that the band-pass filtered flash light (wavelength range between 500 and 600 nm) coincident to the plasmonic wavelength of the copper nanoparticles (Figure 5) in this study,29,30 could induce the photothermal heating of copper nanoparticles during flash light irradiation. Thus, the sintering efficiency of copper nanoink
■
EXPERIMENTAL METHODS
Material Preparation and Fabrication of the Copper Nanoinks. For fabrication of the copper nanoinks, commercially available copper nanoparticles were used (mean diameter,
