Laser-Induced Period Surface Structures to Improve Solderability of Electrical Solder Pads

Abstract

We report on structuring copper representing soldering pads of printed circuit boards by laser-induced periodic surface structures. Femtosecond laser radiation is used to generate low spatial frequency laser-induced surface structures, having a spatial period of 992 nm and a modulation depth of 120 nm, respectively. The slump of screen-printed solder paste is measured to compare the solder coverage on the pads after the solder process on a hot plate. A comparative study of the coverage of solder paste on a fresh polished pad, a pad stored for two weeks, and femtosecond laser-structured pads reveals the improved wettability of structured pads even after storage. In addition, leaded and lead-free solder pads are compared with the particular advantages of the solder-free pad when periodically laser structured. Our findings are attributed to two major effects: namely, the increase of the surface area and the improved surface chemical wettability. Overall, the application of laser-induced periodic surface structures helps to reduce the demand of lead-based solder in the electronic industry and provides a feasible method for a fast and spatial selective way of surface functionalization.

1. Introduction

Electrical connection pads, e.g., metallic solder pads, represent a key part in the packaging of printed circuit boards in electronics [1]. They provide an electrical connection between the electronic circuit and the electronic components. When exposed to ambient atmosphere, the oxidation of such electrical connection pads is a naturally occurring phenomenon [2]. As a result, after transport or storage, electrical connection pads can be covered by a layer of metal oxide, in turn leading to a gradual digression of the solderability of the electrical connection pads [3]. In the worst case, they might be no longer usable and must be disposed with both negative economic and environmental impacts. Therefore, to overcome this deficiency, protective layers are considered. For example, layers of noble metals such as silver or gold are used, covering the electrical connection pads during transportation or storage in order to avoid exposure to ambient atmosphere and hence to prohibit the formation of oxide layers [4]. Nowadays, there are several standards for a protective surface finish of printed circuit boards such as hot air leveling (HAL) [5], electroless nickel immersion gold (ENIG) [6], or organic solderability preservative (OSP) [7]. However, the additional material required for the protective layer increases the production costs and requires dedicated steps and equipment in the manufacturing process of printed circuit boards.

Concurrently, continuous efforts are taken in the electronics industry for improving the solderability of electrical connection pads, since it plays a key role in the durability, performance, and reliability of the resulting electronic devices [8,9]. Driven by the European restriction of hazardous substances directive (RoHS), the demand of lead-free solder material rose and has led to a high demand of environmentally sound soldering joints using tin-based, lead-free solder; however, this is still a challenge for many applications due to, e.g., a higher fusion temperature [10]. Overall, demand continues for technical improvements regarding the solderability of electrical connection pads and their manufacturing without the usage of an additional surface coating. These improvements shall also be applicable for lead-free solders.

A possible method of coating-free surface functionalization is laser-based surface structuring [11]. Especially micro and nanostructures, particularly so-called laser-induced periodic surface structures (LIPSS) generated by ultrashort laser pulses, are highly interesting due to their versatile application [12]. In recent studies, the broad range of surface functionalization based on LIPSS has been demonstrated in fields of structural color [13], renewable energies [14], cell growth [15,16], tribology [17,18] and wettability [19], respectively. In particular, the later application is triggered by both the generation of hierarchic micro- and nanostructures and a surface chemical modification [20], enabling hydrophobic and hydrophilic wetting states.

Chen et al. [8] already demonstrated the possibility of femtosecond-based surface structuring to improve the solderability of Sn-Ag-Cu solder on copper by the laser-based generation of a pillar-like surface with ablation trenches having a width of 10–40 μm, a depth of 20 μm, and a pillar size of 40 μm. These structural features reveal a decreased wetting angle that implies an improved solderability. However, the introduced method using a large trench depth of 20 μm is incompatible for state of the art multilayer PCB having top-copper layer thickness of only several μm. In contrast, the typical modulation depth of laser-induced periodic surface structures is below the applied laser wavelength [21] and therefore appears most suitable for thin solder pads.

Against this background, this study addresses the application of laser-induced periodic surface structures on copper and evaluates the potential influence on the solderability. Therefore, we compare the wetting behavior of leaded and lead-free solder paste for different surface conditions and compare the slump after the solder process.

2. Materials and Methods

Evaluation of the solderability of different surfaces can be performed by different methods, e.g., by the measurement of the 3-point contact line wetting angle between the solder and the substrate. Good wetting is defined as having a wetting angle of substantially less than 90°, whereas an angle larger than 90° implies bad wetting. With increasing wettability, the contact angle decreases to rather small values, which makes it difficult to evaluate such surfaces by the contact angle. Yet another method is the characterization of a material’s tendency to spread the solder after application. The solder spread, also termed slump, can be used as a qualitative and quantitative evaluation of the solderability [22].

Slump stability describes the solder paste’s ability to maintain its structure and shape after printing. Surfaces having a bad solderability will be dewetted during the solder process and will lead to insufficient substrate contact. In addition, the slump of the solder paste should be minimized, as slump creates the risk of forming solder bridges between neighboring pads, creating a short circuit [23]. The general evaluation of the solderability is depicted in Figure 1 and can be divided into 4 major steps: laser structuring, solder past application, solder process, and evaluation.

The laser structuring was performed on commercial available copper sheets with a thickness of 1.5 mm. For a defined initial surface, the pieces were ground with a multi-directional polishing machine (LS3V, REMET, Riale, Italy) with several abrasive strengths (grit size 800–2500) and afterwards polished with suspensions down to a grain size of 1 μm to achieve a reference surface with a roughness $R_a$ of 0.03 μm (measured by laser-scanning microscope VH-X, Keyence, Osaka, Japan). For laser surface processing, we use a micro-machining station (WSMH, Optec, Frameries, Belgium) equipped with a high-power ultrashort pulsed laser (Amphos200, AMPHOS, Herzogenrath, Germany) having a variable pulse duration between 900 fs and 10 ps (FWHM) and a repetition rate of up to 40 MHz. The maximum average power is 200 W with a maximum pulse energy of 1 mJ. For the presented structuring approach, the fundamental emission wavelength of 1030 nm is used. The energy of the laser is adjusted by an external attenuator based on a rotating wave plate and a polarizer. Using an adjustable beam expander telescope, the raw beam diameter is adjusted to a diameter of 5.5 mm (1/e²) (LaserCam-HR, Coherent, Santa Clara, CA, USA). A galvo scanner (Excelliscan 14, Scanlab, Puchheim, Germany) is used in combination with a telecentric lens (f = 163 mm) to focus the beam onto the sample with a calculated spot diameter of 50 μm (1/e²). All given values of the pulse energy and the resulting laser fluence are derived by the laser power measured directly behind the processing optic. To evaluate the laser-based surface modulation, the topography was measured using an atomic force microscope (Dimension Icon, Bruker, Billerica, MA, USA) in PeakForce Tapping Mode.

Via manual stencil printing, the solder paste was applied on the relevant surface. The 0.2 mm thick stencil creates 4 circular shapes of solder paste having a diameter of 6.5 mm. We applied leaded (Sn63/Pb36/Ag2) and lead-free (Sn96.5/AG3/Cu0.5) solder paste with a solder powder size of type 3. The solder process was performed via a flat hot plate and a contact time of 20 s. The temperature was set to 250 °C for both leaded and lead-free solder paste. After the solidification of the solder paste, the measurement of the solder-covered area was performed using an optical 3D profilometer (VR 3200, Keyence, Japan). In addition to solder wettability, the electrical resistance of the solder joints is evaluated using a high-precision resistance measurement system (Milli–TO 3, Fischer Elektronik, Lüdenscheid, Germany) with four-wire measurement Kelvin probes.

3. Results

The first step of our study is the generation of LIPSS on the copper substrate. Referring to our previous studies on LIPSS generation on metals [19,24], the pitch between two consequent pulses is chosen to be 10 μm, i.e., 80% spatial overlap for both the scanning direction and perpendicular thereto. Using a pulse repetition rate of 100 kHz leads to the corresponding scanning speed of 1 m/s. The scanning direction is parallel to the linear polarization direction of the electrical field of the laser beam. A homogenous coverage of LIPSS on the copper surface can be achieved by a laser fluence of 0.8 J/cm${}^2$ and a laser pulse duration of 900 fs. The resulting surface modulation is shown in Figure 2. By the bidirectional scan direction of the hatch, jump delays can almost be neglected, and therefore, a high structuring rate of 10 mm${}^2$/s is achieved. The spatial wavelength of the modulation is 992 nm; i.e., it is slightly lower than the applied laser wavelength of 1030 nm, which was a phenomenon that has been observed before [25]. The orientation of the wavy structures is perpendicular to the polarization of the laser beam. Hence, the structures can be identified as low spatial frequency LIPSS (LSFL) [26]. The mean modulation depth is derived by Figure 2b to be approximately 120 nm.

The first investigation of the potential improvement of solderability is performed on fresh polished and cleaned (ultrasonic and aceton) copper substrate. The generation of LIPSS and the solder test are performed as described above. The solder-covered surface was measured for four samples. The comparison of the solder behavior for the reference (fresh polished) and laser treatment (LIPSS) surface is shown in Figure 3.

The purpose of the initial polishing step is the removal of the natural oxide layer and therefore an improvement in the surface chemical wettability. The comparison in Figure 3 reveals the influence of the topographical modulation generated by the laser treatment. Both surfaces show a very good wettability for leaded and lead free solder. The dashed line marks the initial surface area provided by the stencil aperture of 33.2 mm${}^2$. The lead-free solder enlarges the covered surface after the solder process to $(36.8\pm 1.3)$ mm${}^2$. On the surface having a laser treatment, the covered surface enlarges to an area of $(38.2\pm 1.1)$ mm${}^2$; i.e., the presence of the laser laser-induced micro- and nanostructures enables a higher surface wettability with a 4% larger coverage. A similar effect is observed for the application of leaded solder paste. The application of LIPSS on the copper surface increases the covered surface by 5%. The polished surface leads to a surface area of $(39.3\pm 1.4)$ mm${}^2$, while LIPSS leads to a mean surface area of $(41.3\pm 0.7)$ mm${}^2$. The evaluation of the electrical resistance reveals no significant influence of the surface preparation. Regardless of the surface condition and solder material, the mean contact resistance of the four solder joints is in a range of 14–17 m$\mathsf{\Omega }$.

These findings can be explained by the improved surface wetting on a rough surface described in the Wenzel model [27]. The initial wetting behavior of a smooth surface is governed by the relative surface advance introduced by the roughness. According to our previous studies [20], the surface of LSFL can be described by a $\left|\sin \right(x\left)\right|$ course, and the relative surface advance, considering the modulation period and depth, is calculated using a numerical approach.

$$RSA=\frac{{\int }_0^{\mathsf{\Lambda }}\sqrt{1+{\left(\frac{d_{\pi }cos\left(\frac{\pi x}{\mathsf{\Lambda }}\right)}{\mathsf{\Lambda }}\right)}^2}dx}{\mathsf{\Lambda }}$$

(1)

The measured modulation period $\mathsf{\Lambda }$ of 992 nm and the LSFL depth d of 120 nm lead to a relative surface advance RSA of 4%. The original Wenzel model uses this roughness factor to determine the static contact angle at the 3-point contact line that is in our case inherent with the increased surface coverage of the solder.

Our results show that this obvious small relative surface advance leads to a noticeable improved solderability. Furthermore, the adaption of laser and strategy parameters, e.g., increasing laser fluence, can lead to larger LIPSS modulation depth and the occurrence of additional hierarchical microstructures. For brass, our group achieved a relative surface advance of up to 34% [20].

One possible application scenario is to activate soldering pads with laser structuring directly before the soldering process. The initial materials here would be printed circuit boards, which have already been in storage for a certain period and therefore would exhibit a certain oxide layer on the solder pads surface, thus inhibiting a reliable solder process. In order to investigate the possibility of spatial-selective solder pad activation, the polished copper substrate was stored for 14 days in an air atmosphere at room temperature. The results of this investigations are shown in Figure 4. The oxidation of the reactive copper reduces the wettability of the surface. For solder containing lead, the wetted area corresponds roughly to the size of the stencil but is reduced by approximately 14% as compared to the polished surface, as shown in Figure 3.

For lead-free solder, partial dewetting of the surface takes place after the expiry of storage. As shown in Figure 4b, the solder material slumps from the circular template geometry and thus clearly shows the problem with aged substrate materials. The solder covered area reduces to a value of $(22.8\pm 6.1)$ mm${}^2$.

By applying laser-induced periodic surface structures, the aged surface can be modified in such a way that the increased wettability of freshly polished and additionally laser-structured surfaces is restored for both leaded and lead-free solder. Due to the natural oxide layer on copper under normal environmental conditions being only 10–20 nm thick [28], LIPSS can reveal the underlying copper material. Figure 4a depicts the increase of the covered area in comparison to the aged surface. The laser-based activation leads to a covered surface of $(38.8\pm 0.5)$ mm${}^2$ for leaded and $(40.4\pm 1.4)$ mm${}^2$ for lead-free solder. It is worthwhile to stress that even after two weeks of storage after the laser application, again under ambient conditions, further soldering tests showed that the increased wetting still persists. For leaded solder, the covered surface is $(34.2\pm 1.7)$ mm${}^2$, and for lead-free solder, it is $(32.0\pm 2.8)$ mm${}^2$. Although the wettability is strongly influenced by surface aging, the measurement of the electrical resistance shows no significant deterioration of the electrical properties of the solder joints. For all evaluated surface conditions from Figure 4a, the contact resistance is in the range of 13–19 m$\mathsf{\Omega }$. Thus, we conclude that laser activation by introducing LSFL onto the solder pad using a femtosecond laser poses an effective possibility to improve solderability and storage possibilities, especially for lead-free solder. In turn, this provides an option to foster lead-free solder pads with respect to RoHS guidelines.

4. Conclusions

We have demonstrated the possibility to improve solderability on copper using surface functionalization based on ultrashort laser pulses. In particular, solderability can significantly be improved after storage by ensuring sufficient wettability despite natural oxidation under ambient conditions. We conclude that the application of laser-induced periodic surface structures is a fast, here 10 mm${}^2$/s, one-step method with the potential of a free geometrical adaption of the surface functionalization, as shown in Figure 5.

Solderability was tested for different surface conditions via screen printing using leaded and lead-free solder paste. The measurement of the solder-covered surface implies that the topographical modification by laser-induced periodic surface structures enhances the wettability against a fresh polished smooth surface. This improvement is related to the relative surface advance introduced by the LIPSS and can be derived by the wetting behavior described by the Wenzel model. A possible application of this functionalization is applied as a predecessor process of solder screen printing. The natural grown oxide layer on the copper can be removed, and the topographical modulation in the sub-micrometer scale is applied in one step. Especially, the improvement of soldering lead-free paste is a key benefit of the introduced method and helps to further retain today’s RoHS regulations.