2018 Volume 59 Issue 11 Pages 1811-1816
Our group has previously realized region-selective super-spread wetting through surface fine crevice structures fabricated by laser irradiation or reduction-sintering of oxide powder and joined metals in microscopic regions by super-spread wetting. A new approach suitable for wide-area joining was investigated in this study. Using a metal mesh with high porosity, we produced a new structure termed “interface fine mesh structure” that can promote super-spread wetting. As well as successful super-spread wetting of liquid tin into a copper interface fine mesh structure, copper blocks were also joined by super-spread wetting.
“Super-spread wetting”1–6) is a wetting phenomenon in which a liquid penetrates and spreads into a complicated surface structure containing fine pores by capillary action. Tanaka and co-workers first discovered that super-spread wetting can occur on metal surfaces with fine porous structure formed by the atmospheric oxidation-reduction method.1–3) Using super-spread wetting, they were able to join metals with minimal formation of a solder fillet at the junction. Recently, we have fabricated two types of surface fine crevice structures that can facilitate region-selective super-spread wetting by laser irradiation4–7) and reduction-sintering of oxide powder on the metal surface.8) These two types of surface fine crevice structures that display super-spread wetting may be suitable for micro joining to allow miniaturization of electronic components, such as joining microelectronic wires onto printed circuit boards. With regard to targeted joining over larger areas, it is considered that the methods used to fabricate surface fine crevice structure, i.e., laser irradiation and reduction-sintering of oxide powder, are unsuitable because they require the surface treatment of materials to be joined, which increases the process time and cost with joint area. Therefore, new methods to produce fine structure are needed to achieve super-spread wetting and joining over large areas. One possible approach is to include a porous substance at the interface of the materials to be joined.
In this study, we investigate the use of metal mesh with high porosity as an insert material to obtain fine structure in the joining region, which may lead to region-selective super-spread wetting for wide-area joining. First, we form fine structure at the joining metal interfaces by inserting copper (Cu) mesh between a Cu plate and block. In addition, Cu-to-Cu joining is attempted using the super-spread wetting enabled by the fine structure of the Cu mesh. Furthermore, the transfer route of liquid binder material through the fine structure is investigated.
In the present work, Cu mesh was used to fabricate a new fine structure that can lead to super-spread wetting. The structure obtained by sandwiching Cu mesh between two Cu plates to be joined was determined. The Cu mesh used in this study had a wire diameter of 110 µm, square openings of 144 × 144 µm, and a porosity of 32.1%. Figure 1(a) and (b) show a photograph and scanning electron microscopy (SEM) image of the Cu mesh, respectively. As shown in Fig. 1(b), the Cu mesh has a plain-weave structure in which the vertical and horizontal wires are interdigitated with each other at a constant interval.
(a) Photograph and (b) SEM image of the Cu mesh used in the experiments.
A piece of Cu mesh with dimensions of 10 × 5 × 0.22 mm was ultrasonically cleaned with ethanol. Two Cu plates (99.96% purity) with dimensions of 10 × 5 × 1.5 mm were prepared by polishing their surfaces with 600- and 1000-grit sandpaper and subsequent ultrasonic cleaning in ethanol. As illustrated in Fig. 2, the Cu mesh was sandwiched between the two Cu plates. The sample was fastened together with a clip so that no gaps formed between the mesh and plates. The sample was joined by heating at 773 K for 5 min under a hydrogen (H2; 99.999% purity) gas flow of 20 mL/min.
Illustration of the formation of the interface fine mesh structure.
Figure 3 shows a cross-section of the sample after heat treatment. The wavy line is a cross-section of a horizontal wire and each circle is a cross-section of a vertical wire. It can be seen that fine gaps existed at the contact points between the vertical and horizontal wires and between the two Cu plates and Cu mesh. In addition, the square openings in the mesh were also present as gaps, as shown in Fig. 1(b). That is, three kinds of pores are present in the structure composed of the two Cu plates and Cu mesh. This characteristic structure is termed interface fine mesh structure in this study.
Cross-section image of the interface fine mesh structure.
In this experiment, we investigated Cu–Cu joining using super-spread wetting through the interface fine mesh structure and confirmed that super-spread wetting was possible in the interface fine mesh structure. Tin (Sn; 99.999% purity), that is the main component of the solder alloys was used as the joining material. The melting point of Sn is 505 K.9) As the objects to be joined, a Cu plate (25 × 25 × 1.5 mm) and two Cu blocks (5 × 10 × 1.5 mm) were prepared by polishing and ultrasonic cleaning as described in Section 2.1. Figure 4 displays the appearance of the sample setup for the joining experiment. Cu mesh (20 × 15 × 0.22 mm) was placed on a Cu plate and then Sn wire (0.3010 g) was placed on the center of the Cu mesh. The two Cu blocks were set on both sides of the Sn. The thin Cu mesh used in this study warped easily, so it was difficult to keep the Cu mesh flat and in full contact with the Cu plate. Therefore, it was necessary to compress the sample setup to ensure that the mesh contacted well with the plate and blocks. Pressure was applied to the sample by placing another Cu block (20 × 15 × 10 mm) on top of the two Cu blocks, as shown in Fig. 5. After setting the sample in the furnace, the air in the furnace was removed under the vacuum produced by an oil-sealed rotary pump, and then H2 gas was introduced into the furnace to prevent oxidation of the sample during the experiment. The sample was heated to 703 K for 45 min under an H2 gas flow of 20 mL/min. Once the temperature reached 703 K, the sample was cooled to room temperature.
Photograph of the setup for the joining experiment.
Schematic of the sample setup for the joining experiment.
Figure 6(a) shows the appearance of the sample after joining with Sn. Sn wetted the same region as the Cu mesh. The contact angles of solid Cu and liquid Sn are 37°10) at 573 K and 23°11) at 673 K. Therefore, it is supposed that the contact angle was more than 0° at the present experimental temperature of 703 K, which means that the liquid Sn should not spread out completely on the smooth surface of the Cu plate. In this experiment, the Sn spread out over the entire region of the Cu mesh, which indicates that the interface fine mesh structure occurs super-spread wetting of liquid Sn. Even when the sample was lifted vertically, as shown in Fig. 6(b), the Cu plate and blocks did not separate. Thus, it was found that they were joined well. The cross-section of the sample was examined using scanning electron microscope (SEM) and energy dispersive spectroscopy (EDS). A cross-section of the junction is shown in Fig. 7. The gaps between the Cu plate, block, and mesh were completely filled with Sn and the intermetallic compounds Cu3Sn and Cu6Sn5 were formed at the interface between Sn and Cu. Therefore, it is concluded that the joining occurred with characteristic of the super-spread wetting by the interface fine mesh structure. Also, no fillet was formed at the junction, as depicted in Fig. 7(b). Here, we predict the possibility of super spread wetting with the common Sn-based solder materials by comparing the wettabilities of the solder and Sn on Cu substrate. The contact angles of various Sn-based solder material on Cu substrate have been reported by many researchers.10,12,13) The contact angles of Sn–3.5 mass%Ag and Sn–3.8 mass%Ag–0.7 mass%Cu on solid Cu are respectively 22°10) at 573 K and around 20°12) at 553 K. Sn–37 mass%Pb has 23°13) at 543 K. Thus, Sn-based solder materials, such as Sn–Ag, Sn–Ag–Cu and Sn–Pb alloy, have lower contact angles than the Sn, which means that they have higher wettability on the solid Cu surface than the pure Sn. Hence, it is expected that the Sn-based solder materials cause super-spread wetting through the interface fine mesh structure and the joining using the super-spread wetting of Sn-based solder is possible.
Photographs of (a) the sample after joining and (b) the vertically lifted sample.
Cross-section of the (a) junction and (b) side of junction.
To observe the wetting front of the super-spread wetting, an experiment in which the amount of Sn was decreased to one-third of that used in the initial experiment was conducted by the same procedure as described in Section 3.1. A sample in which the liquid Sn did not spread out over the whole Cu mesh was obtained. The appearance of the wetting front of super-spread wetting is presented in Fig. 8. Figure 9(a) and (b) show surface and cross-section of the wetting front of super-spread wetting. At the wetting front, the Sn penetrated the fine gaps between the horizontal and vertical wires in the mesh and between the Cu plate and mesh. In addition, the Sn in the mesh square openings formed a concave curve because melted Sn preferentially moved along with mesh wire; i.e., the filling of the mesh square openings followed that of fine gaps. It has been reported that super-spread wetting proceeds via gradual wetting of microscopic structure to wetting of macroscopic structure on the surface fine crevice structure produced by laser irradiation.7) It was supposed that the super-spread wetting of the interface fine mesh structure in this study occurs by a similar mechanism to that of the surface fine crevice structure. The gap size of interface fine mesh structure in Fig. 1 is relatively larger than that of surface fine crevice structure formed by laser irradiation.4–7) It is very interesting that the interface fine mesh structure with the opening gap of 144 µm causes the super-spread wetting. The relationship between the possibility of super-spread wetting and the gap size of structure should be clarified from the academic viewpoint on the occurrence of super-spread wetting and, in addition, the engineering viewpoint on the optimal condition for joining with super-spread wetting. On the other hand, it is reported that the conditions for liquid spreading under capillary action in different microfluidic geometries are considered to be different.14) Although the above issue is an essential problem, it is very difficult to evaluate the potential of super-spread wetting due to the complexity of the gap geometries in the structures with several kinds of gaps.
A photograph showing the wetting front of super-spread wetting in the Cu mesh.
(a) Surface and (b) cross-section of the front part of the Sn-wetted area.
A joining experiment using narrow Cu mesh was attempted to confirm that the interface fine mesh structure worked as the route of the liquid Sn during joining. In this experiment, a Cu plate (25 × 25 × 1.5 mm) and Cu block (3 × 10 × 1.5 mm) were used as joining objects, and Sn was used as the liquid. A piece of copper mesh with start and target areas of 5 × 10 mm and a narrow strip with a width of 3 mm between them was prepared. All the components were subjected to ultrasonic cleaning before the experiment. Figure 10 shows a schematic diagram of the experimental setup and side view of the sample. The piece of Cu mesh was placed on the Cu plate. Then, Sn (0.27 g) and the Cu block were placed on the start and target areas of the Cu mesh, respectively. The sample was placed in the furnace, which contained a window that its inside to be observed from the top. The Cu weight used to provide pressure in the previous experiment was not used in this experiment to allow direct observation of the behavior of Sn wetting of the Cu mesh through the window at the top of the furnace. This joining experiment was carried out under the same conditions as the preceding joining experiment. After vacuum evacuation of the furnace, H2 gas was introduced at a constant flow rate of 20 mL/min. The furnace temperature was increased to 703 K for 45 min, and then the furnace was cooled to room temperature. During this experiment, the dynamic wetting behavior of liquid Sn was observed.
Schematic of the sample setup used in the joining experiment.
Figure 11 shows the wetting behavior of liquid Sn from approximately 683 K. It was confirmed that the Sn spread out in every direction within the start area of the Cu mesh. Once the whole of the start area was completely wetted with Sn, the liquid Sn spread to the target area along the narrow path of mesh. Figure 12 shows the appearance of the sample after the joining experiment. The Cu mesh and plate were joined together, but the Cu block was not joined to the mesh. It was observed that the start area and narrow path were completely wetted with Sn, but the target area was not. Here, the amount of Sn required for complete wetting of the 130 mm2 of mesh used in this experiment was calculated. Because the porosity of the Cu mesh is 32.1%, the volume of the gaps in the mesh with a thickness of 0.22 mm is 9.18 mm3. The density of Sn is 6.897 g/cm3 at 704.6 K.15) Thus, the amount of Sn required to completely wet the mesh is 0.0633 g. The actual amount of Sn used in this experiment (0.2754 g) was four times higher than the calculated required amount. As described in Section 3.1, it was extremely difficult to keep the thin metal mesh flat. Therefore, it is considered that the distortion of the mesh generated excessive space in the sample except for the gaps of the interface fine mesh structure created under the appropriate contact conditions. Thus, a large amount of Sn did not wet the target area completely but instead filled the excess space in this experiment.
Wetting behavior of liquid Sn observed from 683 K.
Photograph of the sample after the joining experiment using an interface fine mesh structure with a narrow path.
Therefore, the same joining experiment was conducted again using a different sampling method. Figure 13 shows a schematic diagram of experimental sample setup. Pressure was applied to the sample to prevent the formation of extra gaps that were present in the setup shown in Fig. 10. To apply pressure to the sample, the lower Cu plate and upper block that sandwiched the mesh were fastened together with a clip. In other parts, the mesh was tied to the lower plate with thin wire. Then, the joining experiment was conducted under the conditions described in Section 4.1. The amount of Sn was also 0.27 g. The appearance of the sample after the experiment and a cross-section of the junction are presented in Fig. 14 and 15, respectively. The Sn spread out over the entire region of the Cu mesh, and the Cu block did not detach from the Cu plate when the sample was vertically lifted, which means that the Cu block was joined well to the mesh. The joining was successful because of the super-spread wetting of the interface fine mesh structure by the liquid Sn to fill all the gaps, as shown in Fig. 15. Consequentially, it was confirmed that the interface fine mesh structure can be used as the transfer route of a liquid.
Schematic of the sample setup in the joining experiment using an interface fine mesh structure with a narrow path, clip, and wire.
Photographs of (a) the sample after the joining experiment using the interface fine mesh structure with a narrow path, clip, and wire and (b) the vertically lifted sample.
Cross-section of the junction obtained after the joining experiment using the interface fine mesh structure with a narrow path, clip, and wire.
In this study, we used Cu mesh as a new fine structure to realize super-spread wetting. The joining of Cu plates and blocks using super-spread wetting into the new structure was investigated. The following findings were obtained:
