DAMAGE CONTROL: A Practical Guide to Avoiding Laser Damage to Optics

by James Doty, Ph.D., Alpine Research Optics, 3180 Sterling Circle, Boulder, CO 80301, 303-444-3420, FAX 303-444-1686, E-mail AROcorp@AROcorp.com, http://www.optics.org/arocorp/

Thanks to modern manufacturing practices, thin film coatings for high power laser systems can withstand very high fluences before experiencing catastrophic damage. Yet optics are all too often damaged by laser fluence in real world applications, and end users usually assume that the coatings are not performing to specification which is, in fact, rarely the case. In this article we will examine some of the most common causes of visible and near-IR damage, and will detail some simple precautions that can prevent catastrophic destruction. We will illustrate this discussion with several actual case studies.

Damage Mechanisms

The two most common sources of catastrophic coating damage for high energy pulsed lasers operating in the visible and near infrared are external contamination and "hot spots" in the beam structure. External contamination can be further subdivided into 1) discrete absorption sites such as lint, dust or a material nodule (commonly referred to as spit) and 2) diffuse contamination due to gaseous or molecular contaminants.

For a discrete absorption site, high laser fluence (energy per unit area per pulse) generates a large amount of heat in a very small volume near the surface of the component. If the absorption is sufficient a small explosion is produced that ablates part of the coating, producing catastrophic failure. However, the absorber is also ablated, and what remains is a scattering site with subsurface damage but no absorption. Hence, subsequent pulses cause no further growth of the damage site. Keep in mind that discrete damage sites are usually microscopic (< 50 to 100 µm), whereas diffuse contaminants tend to produce relatively uniform damage over the beam diameter. And when such diffuse contamination is low enough, it is not uncommon to see a nice "beam print" that is characteristic of the distribution of energy within the beam.

"Hot spots" in the beam structure occur when a region of the beam possesses a considerably higher fluence than the nominal value. Most solid-state laser cavities are a thermodynamic nightmare that, under the right conditions, can produce high order transverse modes with very high local fluence. The short pulse duration, high repetition rate and dynamic thermal environment inside the laser can make it particularly difficult to identify when and where this has occurred. Frequently, the peak fluence in a hot spot, which may last for only a pulse or two, can exceed the specified output of the laser by a few orders of magnitude, leading to catastrophic failure of the component.

Identifying the Damage Mechanism

figure 1
200µ x 1mm crescent shaped damage sites in a 2mm diameter beam.
figure 1

To determine which of these mechanisms is responsible for damage, and hopefully prevent its reoccurrence, much can be learned simply by examining the damage site with the naked and eye and through a microscope. One of the first things to check is the relative size and shape of the damaged area compared to that of the beam. For example, if the beam is nominally round but the damage site is crescent shaped (figure 1), the laser has, however briefly, produced one or more high order transverse modes (non-TEM00) resulting in beam instability and hot spots. Hot spots can also be round, but much smaller than the beam diameter (10:1).

When the damaged area and nominal beam are of similar size and shape, then it's possible that a large portion of the optical surface has been exposed to some sort of diffuse contaminant, and damage has occurred due to absorption over the entire beam cross section. It's important to understand just how sensitive high energy laser optics are to contamination. For example, a component designed to achieve a damage threshold in the 20 to 50 J/cm2 range has an intrinsic surface absorptance on the order of parts per billion, and if anything elevates the absorptance even into the parts per million range, catastrophic failure is likely to occur.

figure 2
a. Scalped structure is often indicative of separate but overlapping damage sites.
b. Conclusive evidence is had when the most recent sites can be isolated.
figure 2

Sometimes, however, shape information alone can be misleading, which is why it's often necessary to examine the damage under a microscope. We know of several instances in which a laser was producing hot spots approximately 200 µm in diameter, but each hot spot occurred at a random position within the 2mm aperture of the beam, and after enough overlapping damage sites were accumulated the overall damage appeared to be a reasonably uniform ablation of the coating over the beam diameter, initially misleading us to the conclusion that there was some sort of diffuse contaminant involved. It was not until we examined a new component after only a brief period of use that the "dancing hot spot" became evident (figure 2).

In addition to the shape of the damaged area, another telltale indicator of damage due to hot spots is the timeframe and circumstances under which the damage occurred. Users often relate that, after several hours of operation with no difficulty, they suddenly hear a rapid series of "pops" then discover that a component is damaged. At this point the user must ask himself what changed from one pulse to the next; what in his system changed on a millisecond time frame? The oxides commonly used for coating materials will age only under extreme circumstances such as deep ultraviolet bombardment, but in the visible and near-IR they don't slowly deteriorate to a point where they fail. Almost invariably, the culprit that changed on such a timeframe was the pulse from the laser. If the cavity is on the edge of proper alignment or stability, as thermal conditions in the cavity vary, it is not uncommon to experience a short train of pulses where all of the pulse energy is concentrated into a high order mode. And most deceiving of all, the laser can immediately return to stable operation with excellent beam characteristics. But the damage is done.

Finding and Eliminating the Root Cause

A typical cause of higher order modes is cavity misalignment, which is often introduced when a laser is physically moved between locations. Most commercially available lasers are relatively robust and reasonably immune to the occasional relocation. But lasers that are shared by several groups in large academic or industrial environments are sometimes moved frequently, and are subject to the vagaries of a large group of individuals with varying amounts of training in the care and feeding of the specific laser in use. When higher order modes are observed or suspected, contact the manufacturer to have the alignment of the cavity components checked.

Contamination problems are usually harder to track down but fairly easy to eliminate once identified. There are numerous sources of molecular contamination. Fingerprints, water vapor, airborne oil (from pumps, etc.) and outgassing from adhesives are some of the most commonly encountered. Water vapor (as well as dust) is usually eliminated by physically enclosing the beam path, and filling the enclosure with a positive pressure of filtered, dry nitrogen.

Other sources of contamination can be more insidious and difficult to identify. For example, in amplified systems, parasitic oscillations may produce pulses that are directed outside of the normal beam path, where they can strike component mounts or other equipment. These pulses may then ablate material, such as the black anodization on a mirror mount, or heat plastic or teflon parts, causing them to outgas. This vapor can easily be the source of contamination.

One important indicator that can aid in identifying this particular failure mechanism is the position of the damaged area with respect to the mount. Typically, the contaminant will be more concentrated at the edge of the component where the stray pulse struck the mount, making it a more likely site for damage. Thus, when repeated damage happens in the same position relative to the mount, it points to this type of problem.

Real World Examples

figure 3
Diffuse surface damage due to airborne gaseous contaminants.
figure 3

The constant vigilance that must be maintained in order to avoid contamination from ever-present water vapor is illustrated by the following actual incident. A high power laser user experienced a building power failure during the summer months, and without air conditioning, the laser system was exposed to high humidity for several days. After power was restored, absorbed water vapor was eliminated by operating the air conditioning for two days before powering up the laser. But when the user did so, almost all of the optics in the system were damaged (figure 3). The damage was not a result of water vapor, which was eliminated, but a result of the airborne contaminants (sulphur compounds, traces of tobacco smoke, auto emissions, etc.) carried by the water vapor and left behind when it was "boiled off". Water vapor carries with it a lot of unpleasant "stuff", and if components are exposed to high humidity, they must be cleaned with a methanol wipe prior to use.

In another instance, a laser manufacturer consistently experienced failure on one side of a mounted, symmetric, biconcave lens, and always on the side downstream from the laser, but never on the upstream side. The lens was symmetric, but the mount was not, and an epoxy adhesive was used to join the two; after cementing, the assembly was placed on its side to cure. The subsequent damage was ultimately found to always occur on the side of the lens placed face down during the curing process; gases emitted by the adhesive were being trapped between the lower concave surface and the work surface it was resting on. The top surface, which had proper airflow, was not contaminated. Subsequently, the parts were placed upright (on edge) in a rack and in a laminar flow hood during curing to eliminate the contamination.

figure 4
Discrete damage sites due to dust.
figure 4

One final example shows how misleading first impressions can be. In this case, a near-IR, industrial laser user experienced a gradual degradation in component performance; the nominal beam diameter was 5 mm, and to the naked eye the damage was diffuse and the size of the beam, appearing as a hazy beam print on the component. This lead to the initial hypothesis that the culprit was some diffuse contaminant. However, when viewed under a microscope the damage was clearly a multitude of discrete sites varying in size from 10 to 50 µm (figure 4). Through a process of elimination and a few experiments, it was determined that small dust particles were settling on the surface and being ablated via the mechanism previously described. Each damage site was small, non-progressive, and produced an insignificant loss of energy through scattering. But with time and the accumulation of enough such sites, the losses became significant. Thus, a multitude of minute catastrophic damage events appeared to be a slow decline in performance.

Conclusion

As can be seen from these examples, the true source of laser damage in a particular situation can be very difficult to isolate. Simple visual part inspection can lead to false conclusions. Often, a microscope and the proper lighting may have to be employed to accurately assess the precise nature of the damage. Finally, consult the component manufacturer who usually has a wider base of experience with damage than any single user. And while damage is an unavoidable fact of life when working with high power lasers, through care and understanding it can be minimized.

Cleaning High Power Laser Optics

Oddly enough, the single most common cause of laser damage is using a component without cleaning it, under the assumption that the manufacturer cleaned it properly. The manufacturer did, but under irradiation by the fluences considered here, every cost effective packaging method will contaminate the component to an unacceptable degree. The first time a component is used, clean it with the methanol wipe described in Stage 3 below.

The most important consideration in cleaning optics is to avoid introducing more contaminants than are removed. At ARO, we recommend a three stage process in which subsequent stages are applied only if the previous one fails to eliminate the problem.

Stage 1

Optics should only be handled wearing either powderless latex gloves or powderless finger cots (cleanroom compatible). After gloves are put on, they should be cleaned with methanol to eliminate any contamination from finger oils. Optics should always be handled from the edge, not the surface.

In a clean environment, simply blow any contaminants off the optical surface using an air bulb or ear syringe. We do not recommend using compressed air or nitrogen lines, because these may themselves contain contaminants (propellant, oil, etc.).

Stage 2

Lay a lens tissue on the surface of the component. We recommend using a soft lens tissue, such as ExsorbX 400 (Berkshire Corp.). It is important to note that both sides of this type of lens tissue are not the same; the fibers are looser on one side. The sides can be distinguished by creasing the paper; many more fibers will stand up on one side than on the other. The side on which fewer fibers stand up should always be placed against the optical surface. A drop of solvent is placed on the tissue, wetting the entire surface; the tissue is then slowly pulled off the component. At Alpine Research Optics, we suggest the use of spectroscopic grade (HPLC) methanol, which will leave virtually no residue; we strongly recommend against the use of acetone or isoporoponal.

Stage 3

Fold a piece of lens paper one or more times, again folding it so that the side with fewer fibers sticking up faces out. The reason for folding the tissue is to create a cushion, so that pressure will be applied evenly over the optical surface. Moisten the paper with methanol.

Holding the lens tissue using the index, middle and ring fingers, wipe across the component in one smooth, straight motion, covering its entire surface. We recommend against holding the tissue in forceps, since these will produce an uneven pressure. If the element is too large to be cleaned at once, then clean one half with a "C" shaped motion, then the other half in the opposite direction. When cleaning mounted optics, it may sometimes be useful to fold the lens tissue into a triangle in order to get into the mounted edges. Also, keep in mind that once the tissue has been wetted with methanol, after about twenty seconds it has absorbed too much water vapor from the atmosphere and is useless. And always discard the tissue after a single use.

Figure Captions:

1. 200µ x 1mm crescent shaped damage sites in a 2mm diameter beam.
2. a. Scalped structure is often indicative of separate but overlapping damage sites.
b. Conclusive evidence is had when the most recent sites can be isolated.
3. Diffuse surface damage due to airborne gaseous contaminants.
4. Discrete damage sites due to dust.