Fiber laser technology broadens out

Mid-infrared laser wavelengths open up new processes

Tony Hoult

The telecom crash in the late 1990s can be seen in hindsight as a pivotal event in the history of industrial lasers; from it emerged industrial fiber laser technology. The existence at that time of in-depth scientific and engineering expertise in active fibers and fiber amplifiers, combined with the availability of an extensive toolkit of components, enabled a brisk scaling-up of laser power into the multi-kilowatt regime.

The industry’s shift to fiber laser technology is now challenging well-established lasers in many sectors of the industry. CO₂ and Nd:YAG laser technologies are now almost 50 years old, but the huge success of this new technology comes as no surprise to those who have worked with both the old and the new. The major impacts have been in multi-kilowatt metal processing and in low-power general-purpose laser marking, but there is also rapid progress in other areas. The following presents the emergence of a new process for a new class of mid infrared longer wavelength fiber lasers.

FIGURE 1. IR wavelength ranges and water absorption peak vs. wavelength.1

While the differences between a CO₂ gas laser and a fiber laser are obvious and these two technologies cannot be confused, the difference between a fiber laser and a fiber-delivered laser is not always immediately apparent. The technology change comes from generating the beam within the fiber itself, and the inclusion of other optical fiber components within the same continuous hermetically sealed fiber-optic beam path. This contrasts with fiber-delivered lasers, where the beam is generated by an array of solid-state optical crystals and discrete optical components and is delivered to the workpiece via fiber for only the final part of its journey. The fiber laser is produced by constructing continuous beam paths within fibers; these are very familiar techniques for those in the fiber optics industry. This fit-and-forget process is at the heart of the success of the fiber laser, as it removes the need for maintenance issues associated with other types of industrial lasers.

Fiber lasers in the mid infrared

The infrared portion of the electromagnetic spectrum is usually divided into three regions; the near-, mid-, and far-infrared (FIGURE 1).1 Our concern here is with the higher-energy near-IR regime, approximately 0.8–2.5 μm wavelength. Laser scientists are familiar with CO₂ lasers where electromagnetic bonds between atoms are seen as quantum mechanical springs — for example, the asymmetric vibrational state of the CO₂ molecule is close to the stretching state of the N2 molecule, so energy exchange between the two molecules occurs and lasing in the far IR occurs.

Mid-IR wavelengths are part of the fiber laser story, and thulium was seen very early on as a candidate rare-earth ion dopant for fiber lasers, because the many Tm 3+ transitions allow a wide range of useful wavelengths to be generated using silica-based fibers in the 2 μm spectral region. Strong absorbance by water occurring around this wavelength led to an early interest in thulium fiber lasers for superficial tissue ablation with minimum coagulation depth, a form of “bloodless surgery.” Many other non-materials-processing applications have led to the availability of a range of optics for these wavelengths.

FIGURE 2. Outline of weld volume, 22 layers of 0.1 mm thick virgin LDPE.

It has been known from previous work2 and from spectroscopic data that many polymer matrices absorb more efficiently at this wavelength, but the reason why is not immediately clear. In this mid-IR range, the spectra from even the simplest polymer materials are complicated by numerous vibrational modes and are not simple symmetrical peaks. The wavelength equivalent for a C-H bond absorption peak is typically 3225 nm, well away from the range of the thulium laser. The energy transitions in the stretch vibrational states of these C-H molecular bonds in high-density polyethylene (HDPE), for example, are therefore small compared to the binding energy of electrons in a carbon atom. It appears, however, that the prominent absorption closest to 1940 nm is the first overtone of the prominent fundamental C-H stretching absorption at 1724 nm. Having said that, mid-IR developments are underway and 3225 nm laser wavelengths are now becoming available. This will be an interesting area for future work.

With the availability of much higher average power at these wavelengths and the high water absorption, please ensure relevant safety standards are followed for guidance. Even the phrase ‘less unsafe’ perhaps should not be used!

Thulium fiber lasers for polymer welding

Although 50 W average power thulium fiber lasers have been commercially available for several years, recent developments have pushed average power to >100 W while still maintaining a high-brightness single-mode beam, with a power level and $/W cost appropriate for high-volume manufacturing processes. The existing through-transmission laser welding (TTLW) technique gaining acceptance within industry only allows lap joints with one absorbing and one transmitting component to be welded. Also, as it employs a near-IR laser, it is unable to weld clear-to-clear polymers unless an additive is used, such as Clearweld infrared dye, and this can make the laser process unacceptable.

Preliminary trials

Trials were conducted to confirm the improved absorption levels of a 1940 nm thulium fiber laser beam on many thermoplastics; results on polycarbonate (PC) are reported below. Using a 4.2 mm diameter collimated beam, power was maintained below 5 W to ensure no significant heating or melting.

This data shows that 10%–45% absorption occurs volumetrically in this clear sheet material depending on sample thickness. These results can be compared with those from a commercially available transmission tester that uses a 0.8 mW diode laser source at 850 nm, which showed >92% transmission on all samples.

As heat input to the sample is increased incrementally and by exercising careful control over the temporal and spatial characteristics of the beam, melting occurs in a highly controllable manner in materials up to 6 mm thick. A simple lap joint with very basic clamping was all that was then required to produce optically clear spot welds between two faying surfaces of like thermoplastics. Relative motion between the laser beam and the melt zone produced linear welds, controlling relative speed and power (line energy) penetration in a manner analogous to welding of metals.

FIGURE 3. Material failure surface at interface of 3 mm diameter PC spot weld.

The cross section of the multi-layer joint in low-density polyethylene (LDPE, FIGURE 2) was prepared as a simple way of delineating the melt zone. In polymers this is more difficult than for metals, where conventional metallography can be employed. As with all welds, strength depends largely on the weld area at the interface, and mechanical testing has shown that joint strengths greater than the strength of the parent material can be readily produced. FIGURE 3 shows a spot weld fracture surface in polycarbonate where cohesive material failure has occurred.

FIGURE 4. 4 mm diameter optically clear spot weld through 2 mm thick unmodified polypropylene.

FIGURE 5. < 0.1 mm weld lines in Zeonor cyclic olefin polymer, 2 mm long simulated microfluidic device.

Discussion

FIGURE 2 clearly shows the volumetric absorption of the laser beam in multiple thicknesses of LDPE, which serves two purposes: it shows layers of polyethylene joined together by a high-quality weld, and clearly outlines the melt zone confirming volumetric absorption, as expected from a consideration of Beer’s Law. The classic derivation of this law divides the absorbing sample into thin slices perpendicular to the beam, and tells us that light from each subsequent slice is slightly less intense. For a parallel beam of a specific wavelength of monochromatic radiation passing through a homogeneous solid material, the loss of radiant intensity (ΔI) is proportional to the product of the path length through the material Δx and the initial radiant intensity:3

ΔI = ItΔx

where t is the absorption coefficient and represents the relative loss of radiant intensity per unit path length in the material.

An important contrast exists with 4 mm: the longer-wavelength far-infrared regime is where almost 100% absorption occurs, and the polymer welding process is limited to thin films due to the relatively slow conduction processes involved.

With fine-tuning of the welding process, spot or seam welds with no visible charring or degradation were readily achieved. This simple welding technique has since been applied successfully to a wide range of thicknesses of many polymer materials (FIGURES 4 and 5). For materials ranging from 0.1–3.0 mm thickness, this absorption appears well-suited at this wavelength for welding many optically clear thermoplastics. The use of high-brightness fiber lasers allows long focal length low f-number lenses or even collimated beams to be employed, as large spots and low power density (typically < 500 W/cm²) only are required for polymer welding. As many thermoplastic polymers are well-known for their tendency to distort when welding, this enlarges the operating envelope of the process considerably, and removes the need to focus on a particular plane at the joint interface as has been necessary when lower-brightness lasers are used, so access for tooling is also much improved. Although the Gaussian nature of a single-mode beam may at first glance be considered detrimental, optical techniques for producing top-hat beam shapes are now available for this wavelength4 and may in some circumstances be required. Optically transmissive clamping devices can be used to produce smooth weld surfaces on rigid polymers.

There are many benefits to this process:

Butt and lap joints are possible
Light clamping pressure only is required
Transmissive clamping plate is not always required
No extra absorbers are required
Optically clear defect-free welds are readily obtained
Low-heat-input, sub-0.1 mm wide weld features are achieved
Long focal length lenses or collimated beams rule out access issues.

What may turn out to be most important of all is that this new laser wavelength allows a far greater temporal and spatial control of heat input into polymers than has previously been possible — this may well have some very interesting consequences!

Summary

Fiber laser technology has broadened out into many sectors of the laser materials processing industry. Multi-kilowatt fiber lasers compete with other laser technologies, fiber laser powered marking systems now dominate general purpose marking, new quasi-continuous wave (QCW) fiber lasers replace flash lamp-pumped solid state lasers, low-nanosecond pulsed lasers are now established and sub-nanosecond lasers are under development, and fiber laser components already are widely employed in the pico and femto-second regime. With mid-infrared laser wavelengths opening up new processes such as that reported here, this trend looks set to continue apace. ✺