High-power lasers have become a standard and ubiquitous tool in many industrial applications, in part because of the precisely controlled energy they provide. But "high power" can be a difficult term to parse, and often requires some context to define.
Laser applications are of limited help when it comes to determining high power output thresholds, as laser parameters can range from a 10-watt beam in a 3D printer to a 100-kilowatt beam generated by a satellite laser. What constitutes rich functionality for one application is hardly ever documented in another. When applied to lasers of different wavelengths, when comparing continuous-wave and pulsed operation, and even when comparing solid-state, gas, or doped-crystal sources, "high power" becomes a relative term.
The same beam that can penetrate steel or can relay telecom signals over long distances can also cause serious damage to the sensitive optical components that guide and shape light within the framework of a laser system. "You have to control every last little detail, or those little details will be where the damage starts - and the next thing you know, you've burned up half a million dollars in components," says Tim McComb, manager of global business development at Coherent.
Quality Control
A large number of optical components are needed to ensure that the laser beam achieves the desired shape, size and intensity. In addition to the lenses used to focus and collimate the beam, laser systems typically contain mirrors, polarizers and beam splitters. Each of these components must be precision manufactured and machined, and then treated with specialized coatings to ensure that the final product has the proper light absorption, transmission and reflection characteristics. Without careful monitoring, each stage of fabrication, polishing, coating and testing provides ample opportunity for defects or errors that can lead to system failure.
A component can have defective areas that essentially create weak links in the entire optical assembly. Because the defect absorbs energy that should be transmitted or reflected, the eventual failure of that component can propagate to the rest of the system, says Matthew Dabney, principal laser engineer at Altamont Optics.
The laser's glare can also cause problems through thermally induced deformation of the optics. Even if these effects do not immediately destroy the affected optics, they can cause changes in the refractive index of the material, which can result in distorted or suboptimal laser output. As a result, laser manufacturers need to be aware of a number of considerations when specifying optical components for a particular laser system.
First is their choice of material. Fused silica is a very well-characterized glass with very low absorption and is easy to shape and polish, which often makes it the best choice for many transmissive and reflective optical components.
"We always try to use fused silica when we do high-power applications," said Dirk Hauschild, director of laser optics research and development at Focuslight. "You get the highest level of quality, and the coatings on fused silica have the highest damage threshold."
However, some laser systems require more specialized alternatives. Lens lasers in high-power CO 2 are often made of zinc selenide, which exhibits strong performance in the strong infrared, but can be more difficult to use. Like other optical materials, zinc selenide must be precisely molded and smoothed. Tiny surface defects can lead to performance problems or, worse, localized buildup due to the energy and heat of the laser.Hauschild says this can stress the substrate, which can cause any coating to crack and burn. "For really high-power applications, a single defect can destroy the entire optic."
Eliminating such defects requires meticulous grinding and polishing processes, followed by careful quality control, says Emiliano Ioffe, Ophir's IR process development and engineering manager, whose company typically targets surface roughness values of less than 1 nm on its components, with no scratches or gouges allowed. This is particularly challenging when using non-fused silica materials, and Ioffe says his team had to develop a specialized polishing process for the zinc selenide optics used in the company's CO 2 lasers, especially when preparing materials for aspheric optics that have become increasingly popular lenses.
These perfectly smooth surfaces must then be uniformly coated with specialized coatings that impart the proper reflective or antireflective properties to the components. hauschild says that the coatings are often the weakest link in the design. Because they are so thin, they can fracture, and they can change material properties over time. As a result, poorly selected or applied coatings can negate the hard work that goes into producing the perfect lens or mirror.
In addition to their absorption and reflection properties, coatings must be selected for optimal performance at specific wavelengths. "For ultraviolet, three or four materials are typically used, while for infrared, there is a very different set of three or four materials," says Dabney.
In many cases, components must receive multiple layers of different coatings to improve the desired optical performance, but that improvement may require trade-offs. "You can add more and more layers to improve the reflectivity of the mirror, but when you add layers, they also absorb, so you lose some of the absorbed light," Ioffe says. "There's always a balance between absorption and reflection and transmission."
Multilayer coatings must also be carefully designed to avoid peaks in electric field strength at the interface between layers, which can compromise the integrity of the coating and ultimately lead to component failure.
Finding meaningful fingers
Maintaining a high level of quality control while manufacturing on a commercial scale is never easy. For some basic performance metrics, such as absorption, there are no universal standards that companies can use. "You can't buy a sample with a specific absorption rate as a master calibration system," says Ioffe. "We have developed and built systems in-house to measure absorption, phase shift, reflectance and transmittance at different angles and different polarizations."
The lack of a universal standard is particularly problematic for evaluating the laser-induced damage threshold (LIDT), a metric that describes the maximum energy level a given component can withstand before experiencing measurable damage.
"There are a few ISO [International Organization for Standardization] standards that apply, but these are not sufficient to really create a consistent laser damage threshold test," says Dabney, a member of the American National Standards Institute (ANSI) initiative to develop more detailed test standards.
Hauschild further says that these standards may not be well suited for evaluating the LIDT of optics, as new laser designs continue to push the boundaries of output performance.
LIDT is directly affected by the composition, quality and coatings of the optics themselves, as well as the wavelength and power of the beams they affect. But other factors also come into play. For example, the size and shape of the beam can change the threshold depending on how much energy is distributed over a given surface area of the component. Some of these factors can be modeled mathematically, but an accurate LIDT assessment ultimately requires direct testing of the component itself.
That's where the lack of standardization in LIDT testing comes in. Currently, Dabney says, the ISO standard defines LIDT as "visibly damaged. But that opens the door for notorious manufacturers to sell optics with unrealistically high LIDTs by keeping damage assessment procedures superficial. "You don't get rewarded for looking like you're trying harder-you're actually penalized," he says. As a result, LIDT should be used as an informational guide, not a target for routine operating conditions. Hauschild says his team typically runs its laser systems at power densities well below LIDT to ensure long-term stability.
With great power comes great responsibility, and users must design their laser systems carefully so that a failed component doesn't cause a disaster. "One of the big issues is that if people don't set up their beamline properly, you can get a feedback loop to eliminate the laser," says Dabney. For example, a malfunctioning component could cause the beam to be accidentally reflected back to the light source by a mirror.
"It could cost you $100,000 to laser because there are only $100 parts," Dabney says. Therefore, laser sources should be isolated to prevent such back-reflection events.
Routine maintenance and monitoring of the laser system and its components is also critical. For example, in materials processing applications, the shield windows on high-power lasers tend to accumulate contaminants from the material being processed, and they must be replaced periodically to prevent damage that could allow these same contaminants to seep into the laser itself. In addition to operating the laser below peak power to minimize stress on the system, Hauschild recommends using detectors that can monitor the system temperature or sense evidence of unexpected light scattering that may indicate an impending component failure.
High-power laser systems are not cheap. But efforts to cut corners on expendable components are likely to come back to haunt the end user, who bears the cost of maintaining, repairing and replacing the system.
"If you're driven by power, then you're going to pay for it," says Dabney, but he also points out that this should prompt laser consumers to think twice about how much power they actually need for their particular application. "If you can stay below a certain level, then you can save a little bit," he said.
While certain laser systems will undoubtedly continue to push the limits of power - those designed for defense applications or fusion research, for example - Hauschild said he also sees potential for alternative solutions for some industry users. In some cases, for example, multiple lasers operating at lower peak power can provide similar productivity at a lower cost. "The question is not only whether we should continue to scale up power, but also how we use it in an efficient way," he said.
