1. Introduction

Numerous fabrication technologies for two-dimensional (2D) and three-dimensional (3D) metal microstructures—such as direct ink writing [1–4], electroplating [5–7], laser-induced photoreduction [8–10], laser-induced forward transfer [11–14], and others [15–17]—have been developed in recent years. In particular, two-photon photoreduction (2PR) using a femtosecond pulse laser enables the creation of sub-micron fabrication resolution metal microstructures with a high degree of design flexibility [8–10,16–26]. For example, Tanaka et al. demonstrated the fabrication of 3D metal microstructures with 120 nm resolution by using metal-ion aqueous solutions containing a surfactant as a metal growth inhibitor [25]. In addition, Kang et al. reported a method to fabricate disconnected 3D silver nanostructures for metamaterials with a resolution of less than 100 nm in a gelatin matrix [26]. Furthermore, conductive silver structures have been created for direct writing of silver microstructures. For example, Maruo et al. used a polyvinylpyrrolidone (PVP) film containing silver ions [8], and Nakajima et al. used a mixture of silver ions and polydimethylsiloxane (PDMS) [10]. Owing to the aforementioned features, 2PR has been used for various applications such as electrical wiring [8,10,18], optical devices [19,20], lab-on-chip devices [21,22], and metamaterials [23,24]. However, because of the high fabrication resolution, these methods based on 2PR require a tremendous amount of time to fabricate large structures. Therefore, the size of products that can be made with 2PR is limited, and it is challenging to fabricate metal structures with different scales (e.g., from micro to millimeter-scale) using the same equipment. Therefore, a scalable fabrication method for metallic structures is desired. Moreover, 2PR requires an expensive femtosecond pulse laser as light source, hindering the widespread use of laser direct writing of metal structures. By overcoming the above issues, additive manufacturing for a wide variety of products with various scales, such as sensors and metamaterials, can be established.

In this paper, we first demonstrate multi-scale direct writing of metal structures via single-photon photoreduction using a low-cost 405-nm blue laser. In this method, it is possible to change the fabrication resolution by using objective lenses with different numerical apertures (NAs) depending on the scale of the desired metal structures—from micro to millimeter scale. Although 2PR needs a high-NA objective lens for tightly focusing the femtosecond pulse laser beam, single-photon photoreduction using the blue laser beam allows us to use low-NA objective lens due to high reactivity of single-photon photoreduction. In addition, the use of an inexpensive continuous-wave (CW) semiconductor laser as the light source can reduce equipment cost. In our method, we directly drew silver microstructures in a gelatin matrix containing silver nitrate (AgNO₃) with the laser. In experiments, we investigated the fabrication line width, resistivities of fabricated patterns, and microstructures of fabricated patterns. Finally, we show that micro to millimeter-scale fabrication is possible by changing the NA of the objective lens. The multi-scale direct metal fabrication method demonstrated here would open up new possibilities of conductive metal fabrication and can be applied to various devices such as optical devices, sensors, lab-on-chip, and metamaterials.

2. Experimental methods

2.1 Sample preparation

For the fabrication of metal patterns, gelatin and AgNO₃ dissolved in deionized water were used as the fabrication materials. Gelatin is inexpensive, commercially available. In addition, the sample is also suitable for additive fabrication to the commercial-products, which has complex shapes such as an uneven surface or/and curved surface. The concentration of AgNO₃ (Junsei Chemical Co., Ltd.) was adjusted within the range of 0.2 to 1.2 wt%, and the concentration of gelatin (Morinaga & Co., Ltd.) was kept at 2 wt%. For preparing the sample solution containing 0.8 wt% of AgNO₃, for example, 0.164 g of AgNO₃ was dissolved in 20 ml of deionized water and stirred for 30 min at room temperature (25 °C). Subsequently, 0.412 g of gelatin was added to this solution and stirred in a water bath at 60 to 80 °C until the gelatin was completely dissolved. Then, the prepared solution was dropped onto a cover glass with a pipette and cooled to room temperature to form a gel matrix containing silver ions.

2.2 Fabrication system and procedure

A CW laser with a wavelength of 405 nm (06-MLD 405 nm, Cobolt) was used for the reduction of silver ions. Incident laser power before the objective lens was adjusted from 5 mW to 116 mW with a neutral density filter. The laser beam, which was expanded by passing it through a variable neutral density filter and beam expander, was focused onto the gelatin matrix using an objective lens whose NAs are 0.65 or 0.1 (Plan N, Olympus). By using lenses of different NAs, the curing volume can be changed according to the size of the model. The laser was scanned using a 3D piezoelectric stage when using a high-NA lens and using a galvanometer mirror when using a low-NA lens. The laser beam was scanned at a speed of 1 µm/s to 45 µm/s. Our fabrication system also includes observation optics equipped with a CMOS camera (DCC1545, Thorlabs) to observe fabricated structures during fabrication. A halogen light source with a yellow filter was used to prevent photoreduction of samples, and a longpass filter was used on the laser beam before the imaging lens.

Figure 1 shows the experimental procedure used for fabrication of the silver structures. After the sample had been prepared, the gel matrix was irradiated with the laser, and the desired silver structure was fabricated by scanning the focal point of the laser with a piezoelectric stage or galvanometer mirror. After fabrication, the sample was immersed in deionized water at 60 °C to remove the gelatin matrix, so that only the silver structure remained.

 

Fig. 1. Experimental procedure for direct laser writing of metallic microstructures with gelatin matrix containing Ag ions.

Download Full Size | PDF

2.2 Resistivity measurement

To confirm the conductivity of fabricated microstructure, we measured the current and voltage of a drawn line using the standard four-probe method, as shown in Fig. 2. First, H-shaped silver line patterns were prepared for the experiment. The straight part of the H pattern was scanned 10 times at a pitch of 1 µm to obtain a wide straight line. One side of a gold wire was connected to this H-shaped silver structure with silver paste, and the other side was connected to a measurement system to measure voltage and current. The measurements were carried out by adjusting the voltage with a function generator (WF1974, NF) and measuring the current and voltage with digital multimeters (34401A, Agilent). To calculate resistivity, we measured the feature size of the silver lines using an optical microscope (VHX-5000, Keyence) and a laser scanning optical microscope (VK-X250, Keyence). The height of the fabricated wire was measured using the laser scanning optical microscope, and the average cross-sectional area of the silver wire was determined based on the measured height data. By using the measured resistance, length, and cross-sectional area of each silver line, the resistivities were estimated.

 

Fig. 2. Schematic of four-probe measurement of resistivity.

Download Full Size | PDF

3. Results and discussion

3.1 Fabrication resolution

First, the fabrication line widths with the high-NA lens (0.65) and low-NA lens (0.1) were investigated at different scanning speeds ranging from 1 µm/s to 30 µm/s, and the results are presented in Fig. 3(a) together with the standard deviations of the fabricated linewidths. In these experiments, the concentrations of AgNO₃ and gelatin were 0.8 wt% and 2 wt%, respectively, and the incident laser power was set to 80 mW. The laser power was measured at the point just before objective lens. It was observed that the fabrication line widths decreased with increasing scanning speed. As shown here, various linewidths can be obtained by changing the NA and scanning speed. Typical examples of fabricated line structures are depicted in scanning electron microscope (SEM) images shown in Figs. 3(b) and 3(c), which were obtained using objective lenses with different NAs (0.65 and 0.1, respectively). These results demonstrate that the linewidth could be changed in a wide range from 0.9 µm (NA 0.65, scanning speed 25 µm/s) to 15 µm (NA 0.1, scanning speed 5 µm/s) by changing the NA and scanning speed. The linewidths can be also modified by changing the laser power (data not shown). Such multi-scale fabrication is a unique feature of single-photon photoreduction using a blue laser beam. The multi-scale fabrication is advantageous for large-scale fabrication. In addition, it was observed that these lines were composed of silver particles of various sizes, where the large-sized silver particles tended to precipitate when the scanning speed was slow, and the NA was high. These lines are different from the continuous smooth lines fabricated using 2PR by Maruo and Saeki [8]. However, we observed a large number of silver nanoparticles along the scanned lines. It appears that conductivity of the microstructure is guaranteed by the contacts between these nanoparticles.

 

Fig. 3. Relationship between linewidth and scanning speed for objective lenses with different NAs (0.65 and 0.1) (a). SEM images of fabricated silver lines with high-NA lens (b) and low-NA lens (c). The samples contained 0.8 wt% AgNO₃ and 2 wt% gelatin, the incident laser power was set to 80 mW, and scanning speeds of 1–30 µm/s were used.

Download Full Size | PDF

3.2 Investigation of resistivity

Figure 4(a) shows the relationship between the applied voltage and current of the silver patterns fabricated with samples containing 0.8 wt% AgNO₃ and 2 wt% gelatin at different laser powers. The left and right graphs in Fig. 4(a) were obtained at laser powers of 80 mW and 40 mW, respectively. In both experiments, the high-NA (0.65) objective lens was used and the scanning speed was set to 10 µm/s. The current values in each sample were approximately 10 µA, and the resistivities of the patterns prepared with laser powers of 80 mW and 40 mW were estimated at 1.08 × 10⁻⁴ Ωm and 1.52 × 10⁻³ Ωm, respectively. In this experiment, the optimum concentration of AgNO₃ and the laser power having the highest conductivity were 0.8 wt% and 80 mW, respectively. However, the resistivity of the silver line was quite high compared to that of bulk silver (1.59 × 10⁻⁸ Ωm). Although the resistivity was quite high, conductivity was confirmed in both the silver patterns that were fabricated at 80 mW and 40 mW. These results indicate that even if the resultant silver structure is not a continuous and smooth one, as with 2PR, it still shows conductive properties.

 

Fig. 4. Relationship between current and applied voltage (a). SEM images of silver lines fabricated at different laser power (80 mW and 40 mW) (b). SEM image and EDS map of Ag and C (c). The high-NA (0.65) lens was used with laser powers of 80 mW and 40 mW, the samples contained 0.8 wt% AgNO₃ and 2 wt% gelatin, and the scanning speed was set to 10 µm/s.

Download Full Size | PDF

Figure 4(b) shows the SEM images of the samples fabricated at 80 mW and 40 mW. Particles of various sizes were randomly deposited at 40 mW, whereas a well-organized network structure was partially observed at 80 mW. By observing this network structure in detail, it could be confirmed that nanoparticles are deposited in the network structure. To confirm the constituent elements of the network structure, elemental mapping for silver and carbon was performed by energy dispersive X-ray spectroscopy (EDS) [Fig. 4(c)]. It was confirmed that the network structure contained silver and carbon, both of which had a continuous structure. It can be inferred that this network structure partly responsible for the improvement in conductivity seen in the line patterns produced at 80 mW. It appears that bubbles generated by plasmonic local heating [27] at the focal point due to laser absorption of silver nanoparticles are related to the formation of network structure, but further investigation is necessary to reveal the exact cause. If the generation of such a network structure can be precisely controlled, fabrication of highly conductive microstructures using single-photon photoreduction may be possible.

3.3 Multi-scale drawing

Finally, in order to demonstrate multi-scale drawing, we fabricated a complicated pattern using two objective lenses having different NAs (0.65 and 0.1). In this experiment, samples containing 0.8 wt% AgNO₃ and 2 wt% gelatin were used. The scanning speeds using high-NA and low-NA lenses were set to 10 µm/s and 45 µm/s, respectively. Figures 5(a) and 5(b) show the resultant silver patterns made using the high- and low-NA lenses, respectively. Figure 5(a) was produced by scanning the lines were 10 times at a pitch of 1 µm (same as H pattern) to obtain a wide linewidth. Meanwhile, Fig. 5(b) was generated by scanning the lines 10 times at a pitch of 10 µm, and the scanning was conducted twice along the same trace to obtain a wide linewidth and thick line. The electrical continuity was simply confirmed in both the structures using a tester. As can be seen from Fig. 5, we succeeded in fabricating silver metal structures with various scales by changing NA of the objective lens. The fabrication time for these models was approximately 0.8 h using the high-NA lens and approximately 6.6 h using the low-NA lens. It should be noted that single-photon photoreduction is particularly suitable for large and high-definition metal structures, for which 2PR is unsuitable. Metallic luster was also confirmed in a large-scale pattern fabricated with the low-NA lens.

 

Fig. 5. Multi-scale fabrication of complicated patterns using high-NA lens (a) and low-NA lens (b). The high-NA (0.65) lens was used with laser power of 80 mW and scanning speed of 10 µm/s. The low-NA (0.1) lens was used with laser power of 116 mW and scanning speed of 45 µm/s. The samples containing 0.8 wt% AgNO₃ and 2 wt% gelatin were used in both experiments.

Download Full Size | PDF

4. Conclusion

We proposed and successfully demonstrated a novel method for multi-scale fabrication of metal structures using a 405-nm blue laser. We fabricated a silver microstructure by using a gelatin matrix containing silver ions through single-photon photoreduction. We found that the resultant silver structure fabricated at high laser power formed a well-organized network structure, which is partly responsible for improving the microstructure’s electrical conductivity. Finally, we demonstrated that multi-scale drawing (from micro to millimeter-scale) is possible by using objective lenses with different NAs. In the future, we will investigate the reasons for the formation of network structures and further processing conditions to improve conductivity. This cost-effective metal fabrication technique can open up many new possibilities for conductive metal direct writing.