Joining the un-joinable with ultrafast light

Lasers have, and continue to be, an important tool in the manufacturing sector. From the cutting of sheet metal for car doors, to microscopic sculpting of the smallest components, the range of laser applications is staggering. At the Engineering and Physical Science Research Council (EPSRC) Centre for Innovative Manufacturing in Laser-Based Production Processes (CIM-Laser),¹ cutting edge laser development and laser processing techniques are combined for the introduction of new manufacturing capabilities in the UK. A key area of interest is the manufacture of complex and high-precision systems in which a range of specialized materials and material combinations are used. It can often be challenging, however, to achieve direct and robust joining of highly dissimilar materials (e.g., glass to metal).

Traditionally, highly different materials have been bonded by using a thin, intermediate layer (e.g., solder or adhesive) between the parts that are to be joined.^{2, 3} However, issues that arise with regard to inaccuracy, creep, or outgassing limit the use of such interlayers. In particular, there is a significant problem in joining optical materials (such as glasses or crystals) to structural materials (e.g., metals).

As part of our work at CIM-Laser,⁴ we are therefore developing a new technique, in which we use ultrashort pulsed lasers for the direct bonding of problematic materials. Such an approach, i.e., ultrafast laser welding of highly dissimilar (glass to metal) materials, was first reported with the use of femtosecond lasers in 2005.⁵ More recent demonstrations, with picosecond pulses,^{6, 7} have made the process more useful for practical applications. In this method, a train of ultrafast laser pulses provides highly localized deposition of energy at the joint interface, through a combination of linear and non-linear processes. At the same time, the extent of the heat-affected zone (HAZ) can be controlled (see Figure 1). This unique combination means that we are able to bond materials with significantly different (up to 100 times) thermal expansion coefficients in a process that is essentially cold (the HAZ is restricted to a few 100 micrometers around the weld).⁸

Figure 1. Schematic illustration of the laser microwelding principle. In this case, a glass and a metal are being welded together. The laser is focused onto the interface through a top glass surface that must be transparent. As the laser is translated along the interface, a weld—surrounded by a narrow heat-affected zone (HAZ)—is created.

In principle, ultrafast laser welding is extremely straightforward. By placing the focus of the laser at, or around, the interface of the two components, the temperature of the materials can be increased in a highly localized fashion. This results in the rapid creation of a plasma, which is formed from both materials (see Figure 1). This plasma is confined to the interface region and will thus mix, cool, and recombine to form a true bond between the two materials. The two main difficulties associated with the process are ensuring a balance in the absorption of the incident radiation between the two materials, and making sure that the plasma remains confined for long enough to form a bond.

The stability of the created plasma is generally linked to the size of the gap between the two materials. If the gap is too large, ablation, rather than welding, will result. It is usually thought, therefore, that the gap between the two materials must be sufficiently small to prevent the plasma from escaping, and in most cases, optical contact (<1μm) has been regarded as a requirement.⁹ This extreme requirement represents a significant limitation in terms of material preparation. Nevertheless, in our work, we have demonstrated that such strict specifications are not necessary. Instead, we have shown that manipulation of the melt surrounding the plasma can help its confinement (and thus allow material preparation requirements to be relaxed). This effect is particularly pronounced when transparent materials are welded together because the melt can be generated below the interface and, through flow, fill any gap (and thus contain the plasma), as illustrated in Figure 2.¹⁰

Figure 2. Schematic cross section through an ultrafast microweld produced in a transparent material. By carefully positioning the focal point of the laser below the material interface, the process can generate a HAZ (melt) that will expand upward to fill any gap between the material. A more detailed description of this process is provided elsewhere.¹⁰

To ensure simultaneous plasma formation in both materials for welding glasses to metals (see Figure 3), it is critical to position the focal volume in the correct plane. It is not a trivial matter, however, to determine the correct focal plane. In our work, we have thus concentrated on parameter mapping and characterization of the quality of the produced welds. In particular, we use shear strength as a metric for assessing the quality of the bonds (see Figure 4). These results indicate that the strengths of the welds are greater than those of commonly used adhesive materials. Furthermore, fully characterizing the welds is essential for gaining a full understanding of the ultrafast welding technique and for producing a reliable, industrially relevant technique.

Figure 3. Micrograph of an example weld between a glass (BK7) and a metal (aluminum: Al 6082). The weld is in the shape of a spiral, with a pitch of 156μm and an outer diameter of 2.5mm.

Figure 4. Weibull plot of the shear strength measured from 20 Al 6082–BK7 bonds (similar to the weld shown in Figure 3). The plot indicates that the failure mechanism of the bond is complex, with more than one potential cause.

In summary, we have demonstrated a new technique for bonding highly dissimilar materials in the manufacturing of high-value components. Our results to date illustrate the potential capability of our approach. For example, we can produce bonds that have strengths in excess of currently used adhesives. From our proof-of-principle demonstrations we have sought to increase understanding of this extremely powerful technique. In our future work we will concentrate on optimizing the parameters for stronger bonds, and on increasing yield and repeatability with the goal of producing a truly industrially ready process.

This work was funded by EPSRC, through the EPSRC Centre for Innovative Manufacturing in Laser-Based Production Processes (grant EP/K030884/1). Part of the reported work has also been funded by our industrial partner Leonardo Airborne & Space Systems.