Advances in laser processing expand laser markets and production techniques
by Matt Henry
The laser industry has been punctuated by step changes, where existing industrial processes have been replaced by a laser process. This has expanded the total available market for lasers and fundamentally changed a number of industries’ approaches to production techniques. In addition, step changes have resulted when a new class of laser technology has powered an innovative approach to an emerging market.
Figure 1. Lithography - multiple steps vs. Rapid Laser Patterning (RLP). Significant capital cost savings and significant operating costs savings with RLP.
The unique nature of Q-switched diode-pumped solid-state lasers (DPSSL) has been part of the change in the new, and growing, market for Rapid Laser Processing (RLP) of flat panel displays. RLP of Transparent Conductive Oxides (TCOs) replaces the need for expensive production equipment, streamlining production and reducing capital and running costs by up to 50 percent.
RLP provides an opportunity to eliminate the use of large size photomasks and associated lithography tools in the patterning of ITO coatings (and other types of thin films both conductive and non-conductive) in the production of plasma display panels (PDPs). Processing speeds and quality exceed or match conventional lithographic processes, eliminating the need for multi-million dollar area steppers and patterning equipment. Furthermore, the removal of tooling allows production changes to be implemented straight from the CAD systems to the production tool, eliminating reliance on complex masking technology and streamlining production (see Figure 1).
Major changes are not achieved without significant efforts into research, development, and testing. Engineers at Powerlase Ltd. have developed diode-pumped, high average power, nanosecond pulsed, kilohertz repetition rate infrared lasers for RLP. These lasers are now available at average powers nearing one kilowatt, allowing productivity to take this process from the laboratory into industry. This technology is now in place at multiple global sites in full 24/7 mass production of PDPs.
Figure 2. Typical PDP architecture
PDPs require positional accuracy on the order of 5mm -within the capabilities of modern laser scanners, and feature sizes are of the order of 1x1 mm with a resolution of 10mm-all ideally suited to RLP. Given the large areas of PDPs, 429 inches and above, a high-speed manufacturing technique is required (see Figure 2).
Due to the commercial drive for alternatives to lithography, RLP has been investigated for virtually all commercially available short-pulse lasers (nanosecond or below) for processing ITO-ranging in wavelength from the Infrared (IR) to the Deep Ultraviolet (DUV).
Excimer lasers offer nanosecond pulses at UV wavelengths and are widely used for precision micro-fabrication. It is known that ITO and other TCOs can be successfully removed from glass using KrF excimer lasers at 248 nm wavelength. However, this requires precise process control to selectively remove the TCO as the excimer pulse can etch and damage the glass beneath. Furthermore, excimer lasers are not favored by industry due to high cost of ownership and safety issues stemming from the use of corrosive halogen gases.
Ultrafast lasers operating in the picosecond and femtosecond regime have also been investigated for a variety of thin films on glass. High-quality thin film removal has been demonstrated without glass damage for both solar cell and FPD applications. However, it was found that ultrafast lasers have relatively low pulse energies-on the order of 1mJ. Thus to achieve thin film removal they must be focused to fine spot sizes on the order of 10mm to achieve sufficient energy density (fluence). This renders them unsuitable for creating large area TCO structures at the commercial rates required for large area FPD.
Q-switched DPSS lasers offer an answer to the successful introduction of RLP in the flat panel PDP market. Such lasers offer average power levels >800W, at kilohertz repetition rates, with nanosecond pulse durations and pulse energies of >100mJ. Such high pulse energies allow the use of beam delivery techniques more commonly associated with excimer lasers. The beam is homogenized, imaged onto a mask, and re-imaged onto the substrate. The high pulse energies allow sufficient energy density to ablate large pixels (>1mm2) with a single pulse. Kilohertz repetition rates mean that thousands of pixels can be ablated per second-and by ‘stitching’ the pixels together large areas of active ITO electrode structure can be created rapidly. To further improve throughput in industrial systems, multiple lasers may be employed to achieve commercial throughputs that exceed those of conventional lithography (see Figure 3).
Figure 5. SEM image
To demonstrate that this technique was ready for mass production, Powerlase has engaged in a detailed study of the ITO removal process. Employing three analytical techniques, optical microscopy, SEM, and AFM, Powerlase has proved that RLP of ITO on glass can completely remove all the ITO without any substrate damage, and an optimum ablation threshold was identified on example substrates. Additionally, it was found that the process is tolerant of energy variation and has a wide operating window. It is also clear that the quality of the process is very high, with sub-micron edge definition and no dross. It was also noted that a ridge is created around the edge of the ablated region, but given that this ridge is <100nm of the unprocessed ITO it is concluded that this will not affect the function of a PDP in practice.
It was also demonstrated that using AFM image (see Figure 4) and SEM image (see Figure 5) achieved electrical isolation at a significantly lower fluence than for full ITO removal, but that higher fluence is required for complete removal. Combining all this corroboratory data, Powerlase showed that ITO is completely removed by the laser direct write process, and that high process throughputs could be achieved with its Starlase range of lasers. This conclusion is supported by the high number of PDP sets now on the market that are manufactured using this technique. This process has no impact on the end product, and allows producers to offer sets to a wider market at lower prices (see table).
Commercially, compared to wet-etch lithography, RLP has been adopted as the technique of choice for mass production in PDP manufacture. This technique can also be employed in a number of other FPD and thin film applications. Powerlase is now looking to deploy this technology in a number of other key industrial areas, helping to expand the total laser world market and promote lasers for new industrial uses and processes.
Matt Henry is applications manager for Powerlase, www.powerlase.com.