Caption: Copper has 13 times higher absorption of blue laser light compared to infrared
Fibre lasers are dominating the material processing market right now, but many metals only absorb partially in the 1µm wavelength emitted by solid-state lasers. Now, high-power diode laser systems are beginning to be released based on blue laser diode technology, at which wavelength pretty much all metals have higher absorption.
Industrial blue laser diode systems from US firm Nuburu and Japanese company Shimadzu were exhibited for the first time at SPIE Photonics West in San Francisco early in the year – Nuburu was a finalist for a Prism Award at the event. In addition, Laserline presented work at the conference from the German-funded Blaulas project, which is developing high-power CW blue laser diode bars.
Gallium nitride, the material blue laser diodes are made from, is manufactured in large volumes for displays and illumination, but only now has it matured to a point where it can be used for material processing, according to Jean-Michel Pelaprat, co-founder of Nuburu.
Dr Christoph Ullmann, managing director of Laserline, explained: ‘In the past, the output power of blue diodes was not at a level where you could build lasers for material processing. Today, with the 50W bar [from the Blaulas project] we are now at a level where it’s possible to build systems with some hundred watts. With this, you can start material processing.’
At Photonics West, Nuburu showed a 150W blue diode laser module, the AO-150, which it launched in 2017, while Shimadzu was exhibiting 100W modules. Nuburu also gave a scientific presentation during the high-power laser session at the show, reporting a prototype device with 730W of blue laser power. Nuburu intends to launch a 500W laser late this year, and the company has a roadmap to go to multi-kilowatt power levels, Pelaprat said.
Nuburu’s technology combines single gallium nitride diodes from Osram Opto Semiconductors with micro lenses; multiple diode beams are collimated into a single beam, which is then coupled inside an optical fibre. ‘The technology around the micro optics, the fast axis and slow axis collimators, and the macro optics to combine this into a fibre, is where a large part of the knowhow of Nuburu resides,’ Pelaprat said.
A blue diode bar emitting slightly above threshold. Credit: Coherent-Dilas
Meanwhile, Laserline presented work at Photonics West on a 700W fibre-coupled CW diode laser, operating at 450nm with 60mm mrad beam quality. These are results from the first two years of the three-year Blaulas project, which aims to build a kilowatt blue diode laser. Osram, Coherent-Dilas and the Max Born Institute are all partners.
Benefits of blue
‘All metals have higher absorption of blue wavelengths [than the infrared],’ commented Pelaprat. ‘The value exists for all metals. It’s a very large market.’
Yellow metals benefit most from processing with blue laser diodes because of higher absorption at 450nm versus absorption at 1µm – gold is more than 100 times more absorbing in the blue than the infrared. But processing copper is where the blue laser could come into its own. The metal has 65 per cent absorption in blue wavelengths versus 5 per cent in the infrared, 13 times higher absorption. Aluminium is three times better absorbing – 15 per cent in the blue compared to 5 per cent for infrared – while stainless steel is only 50 per cent higher for blue compared to infrared.
‘At low power, some of our customers using the laser for copper welding applications have reported a speed increase of eight to ten times versus infrared lasers,’ reported Pelaprat. Most applications that would use blue diode lasers are for welding at the moment, largely because of the lower beam quality compared to fibre or disk lasers.
Welding thin copper foils in batteries is one promising application for blue diode lasers. Credit: Nuburu
‘In general, a 200-250W blue laser is equivalent to a 1.2-1.5kW infrared laser in terms of welding speed for copper,’ Pelaprat continued. ‘It is about a factor of eight better, but it depends upon the metal you are joining. It requires almost an order of magnitude less power to perform the same application. If you look at copper applications, people are using 6-8kW and even 10kW lasers to join thick metals. We envisage needing a 1-2kW blue laser to process thick metals.’
Pelaprat noted that the blue laser’s speed improvement and the higher quality welds it produces when machining thin metals diminishes for thick metals. ‘It won’t be eight times faster when welding 4mm-thick copper, as is the case for joining copper 0.5mm thick. However, it will still be faster,’ he said. ‘We don’t know how fast a blue laser can weld thicker metals because we haven’t tried it yet, but we anticipate this to be lower – it’ll still be more efficient absorption, but it will be less of a speed advantage at the same power level.’
Nuburu is focusing on copper processing at the moment. The first application it is targeting is for welding copper foils in lithium ion batteries. Inside a battery cell for a car, there are multiple copper foils on the cathode side interleaved with aluminium foils on the anode side. It is crucial to join the copper foils with a high-quality weld.
‘We have demonstrated extremely high quality – essentially defect-free – welds of 40 foils at 2m/min,’ Pelaprat stated. ‘That is not possible with infrared lasers; fibre lasers cut the foil.’
Joining copper foils is carried out today mainly by ultrasonic welding. This is a contact tool with a complex geometry, so only a certain type of joining can be achieved. Ultrasonic welding also generates particles, which can contaminate the battery cell. ‘People are looking for a solution that will give defect-free, contactless welding,’ Pelaprat said.
A 135W prototype blue diode laser developed by Coherent-Dilas
Using blue laser light creates a defect-free weld with low thermal impedance, which reduces the temperature rise inside the weld.
‘Every application working with copper is of interest,’ noted Ullmann at Laserline. ‘There is a lot of activity around fabricating batteries for electromobility. The question is how to join copper with other metals, or, for example, weld 10µm copper foils, which is very difficult to do at the moment with an infrared laser. With a blue diode laser you can do heat conduction welding and get a very nice join. Also, it’s difficult to weld 300µm copper foils with infrared lasers. This can be done partly with a pulsed laser, but at limited speed. With the blue diode laser you can reach a few metres per minute welding speed.’
The blue laser also minimises the intermetallic when joining dissimilar metals, such as copper to aluminium, which increases the mechanical strength of the weld. The intermetallic is a brittle alloy generated at the junction between copper and aluminium when welding with an infrared laser.
There are different approaches to building high-power direct diode lasers, according to Ullmann. One is to use single emitters, which in the near infrared might emit 10-15W of laser light, and combined will give a few hundred watts of power. The other approach is a diode bar, typically a 10mm-long chip that produces around 200-250W output power in the infrared. The same methods can be used with blue diode lasers: blue single emitters have an output power of 3-5W, while the bar concept can produce 50W per bar, Ullmann said.
Laserline bases its technology on stacked diode bars with actively cooled heat sinks. Combining eight 50W bars, for example, will give 400W output power from a stack. The laser light is combined by polarisation coupling, among other methods.
‘With 20 years’ experience of working with infrared diode lasers, we can transfer a lot of technology knowhow to working with blue diodes,’ Ullmann said. ‘There are some aspects that are new, but beam combining, stack combining and beam shaping are very similar to what we have done with infrared diode lasers.’
Coherent-Dilas is working with its T-bars as a partner in the Blaulas project. The firm’s core competence is building high-power diode lasers from monolithic multi-emitter bars. Their expertise is in bar mounting, which involves proper cooling designs and material selection, beam shaping for collimation, and coupling the light into a fibre.
Schematic of a Coherent-Dilas T-bar plate
Coherent-Dilas’ T-bars are the building block for pumping multi-kilowatt industrial fibre lasers, and are manufactured in high volume, automated production. ‘Coherent-Dilas can use its manufacturing capability and even the platform for the fibre laser pump, for fabricating blue diode bars in the future,’ stated Dr Florian Lenhardt, product line manager for high power diode lasers at Coherent-Dilas. ‘The key is a very compact, modular and cost-efficient design, which is power scalable.’
Lenhardt added that one of the targets for the Blaulas project is to achieve homogeneous blue diode wafers, so bars can be cleaved from the material, and not only single emitters.
Ullmann stated that power and beam quality are important to make blue diode lasers commercially attractive. ‘Applications exist where you can work with a few hundred watts with 30-60mm mrad beam quality. These need to be verified,’ he said. ‘But if you have higher output power you will open more applications. It’s always a question of which application cannot be addressed at the moment with other technologies. Customers are willing to start with lower power and slower speed with the understanding that power levels will increase.’
Nuburu’s direct diode blue laser has a brightness of 20mm mrad, which is not an issue for welding, noted Pelaprat. The beam quality, however, means that ‘this particular laser, at this stage, is not yet ready for cutting applications or some of the powder bed applications,’ he said. The AO-150 and the 500W version will be suitable for welding, brazing and laser metal deposition.
The wall plug efficiency of Nuburu’s laser is a little better than half of that of a fibre laser. Gallium nitride, as a semiconductor, has an efficiency of roughly half of the infrared diode semiconductor, but it is direct generation of blue light from electricity. However, ‘the application efficiency is probably going to be four or ten times better,’ noted Pelaprat. ‘So, in terms of net energy, we are much better than an infrared laser. You lose half on one side, but you gain a factor of four to ten on the other side.’