High Laser-Induced-Damage Threshold Optics
Trey Turner, REO Inc.
Despite nearly 50 years of study, laser damage continues to be a
topic of some confusion for both optics manufacturers and consumers. Laser optics
purchasers need to know the primary causes of damage and the steps that fabricators
typically take to minimize them. They must engage in a meaningful dialogue with
potential vendors to understand what questions to ask when qualifying a supplier.
Laser damage regimes
Damage is a problem in virtually every wavelength regime in which
lasers operate. However, the damage mechanisms of lasers operating in the visible
and near-visible spectra are distinct from those of lasers operating in the far-infrared
and deep-ultraviolet. For brevity, this article will concentrate on lasers operating
in the visible, near-ultraviolet and near-infrared. Within this range, it is useful
to further categorize lasers according to their temporal characteristics: namely,
ultrafast, nanosecond and continuous wave.
An operator loads an optic into REO’s
automated damage test system, which enables testing at 1064 nm with either 3- or
20-ns pulse durations at pulse energies of up to 50 J/cm2. Images courtesy of REO
Inc.
For all these laser types, damage is usually defect-related; however,
the way in which damage occurs depends upon the operating regime. Ultrafast (pico-
and femtosecond) lasers typically produce high peak powers and low energy per pulse;
furthermore, the time between pulses is long compared to the pulse duration itself.
Consequently, the primary cause of damage in ultrafast optics is usually related
to instantaneous field strength, rather than to total pulse energy. Specifically,
the electric fields present in ultrafast pulses can be sufficiently high to cause
dielectric breakdown of a defect or even the thin-film materials, resulting in arcing
and damage. Generally, there isn’t adequate energy in an individual ultrafast
laser pulse to heat a defect enough to cause purely thermal damage, and there usually
is plenty of time between pulses for heat to be conducted out of the defect into
surrounding material, so there is no heat buildup over time.
Nanosecond-regime lasers typically operate by Q-switching and
produce pulse durations in the 100-ps to 100-ns range, with repetition rates in
the tens of hertz to hundreds of kilohertz. Under these circumstances, damage usually
occurs by absorption, heating and, ultimately, vaporization of defects and the surrounding
material. As a result of complex thermal and mechanical interactions between the
defect materials and the coating materials, coating materials with the lowest absorption
do not necessarily yield thin films with the highest damage threshold.
One additional complexity of specifying and producing high laser-induced-damage
threshold optics is that defects that may induce laser damage can be very small.
Depending upon its composition and absorption characteristics, a defect of only
a few tens of nanometers – invisible to standard inspection equipment –
can be problematic.
Lasers can cause a lot of damage to optical
coatings, and buyers of optics would do well to understand laser-induced damage
and how to avoid it. Images courtesy of Alex Martin, REO Inc.
Thus, for nanosecond-regime lasers, damage is dependent upon total
laser pulse energy, material absorption and thermomechanical characteristics of
both the coating materials and defects, and defect distribution. Repetition rate
also becomes significant when it is shorter than the time necessary for heat to
be conducted away from absorbing defects.
For lasers that are CW or quasi-CW, the absorption characteristics
of defects and coating materials are the main determining factor in damage. Because
there is no thermal recovery period, damage initiated at a small defect by a high
CW fluence will often spread to include surrounding areas until the entire region
within the high-fluence beam area has been damaged catastrophically.
(Left) CW laser damage caused by a 1-MW/cm2 beam at 1030 nm. The damage site propagates
to cover the entire size of the beam. (Right) Numerous pulsed-laser damage sites,
caused by a 27-J/cm2 pulse at 1064 nm. Damage does not propagate, and each site
is relatively small.
Fabrication issues
A variety of defects that can cause laser damage are introduced
at each stage of the manufacturing process. Thus, it is worthwhile to review in
detail the sources of the defects, as well as the fabrication, cleaning and coating
methods that eliminate them.
One problem that occurs during component fabrication is subsurface
damage. This consists of fractures and scratches caused by the various grinding
and polishing processes, which become partially or totally hidden during subsequent
fabrication steps. It has been shown that a thin layer of the substrate material
– the Beilby layer – can reflow while the piece is being polished, sealing
these defects below the final surface. Subsurface damage can reduce the damage threshold
by providing a place for light-absorbing contaminants to reside, by allowing atoms
at or near the fractures to be more easily ionized, and by causing local intense
modulations in the electromagnetic field.
The fabricator can take a couple of approaches to minimize subsurface
damage. First, grinding and polishing should occur in a series of small steps in
which the size of the grit used is decreased gradually, rather than in large increments.
Specifically, each polishing step should remove enough material to eliminate the
fractures from the previous step without itself producing any that are larger than
those from the previous. Superpolishing is also a useful technique for fabricating
high-damage-threshold optics. The term “superpolishing” broadly refers
to any method that yields very low surface roughness, <1 Å rms.
High laser-induced-damage threshold optics
are produced over a wide range of sizes and spectral ranges.
Cleaning of optical surfaces, particularly just prior to coating,
is a critical step. The goal is to remove contaminants from the optical surface
before they are sealed in by the coating. The types of contaminants found in a typical
fabrication environment can generally be classified as either particulates or
organics.
Particulates are things like residue from the polishing process
and dust; these are usually removed with detergent cleaning chemistries and ultrasonic
techniques. Organics are often airborne contaminants or surface residuals from earlier
cleaning and blocking processes. Typical examples are human sweat, finger oils,
blocking wax, adhesives and organic cleaning solvents. These are usually removed
with other organic solvents and detergents. The takeaway here is that a variety
of cleaning techniques must be employed to thoroughly prepare a surface for coating.
Control of the production environment to eliminate sources of
contamination is also critical. The best practice is for the entire cleaning and
coating process to occur in a cleanroom, rather than under a laminar flow hood.
Furthermore, it is highly advantageous to work with a vendor that can fabricate
and coat in a single facility, eliminating the potential contamination from packing,
shipping and unpacking steps.
Coating considerations
Currently, high-damage-threshold coatings are made using a variety
of deposition processes, including various evaporative techniques and sputtering
methods. Most coating manufacturers tend to employ a single method and therefore
try to convince customers that their chosen approach is uniquely suited to delivering
high-damage-threshold thin films. At REO, we have ion beam sputtering (IBS), ion-assisted
deposition processes and traditional evaporative methods at our disposal, so we
can afford to be unprejudiced in evaluating various coating techniques.
Most of REO’s high-damage-threshold coatings are produced
using IBS because it yields fully densified films that, after they are formed, are
essentially impervious to the absorption of contaminants. Also, a properly managed
IBS process creates films with relatively few internal defects. Thus, IBS films
will generally not damage readily. However, because they do exhibit relatively high
internal stress, once damage is initiated in a dense film, it is likely to be catastrophic.
In contrast, the coatings produced with evaporative techniques
are relatively porous and contain low internal stress, which makes damage sites
grow more slowly than in a denser film. But this porosity also allows higher levels
of particulate contamination within the film, as well as entry of water or volatile
organics. A porous film can absorb airborne molecular contaminants even during use.
Furthermore, during the evaporative process itself, liquefied coating material can
spatter onto the film, producing elements that aren’t stoichiometric and
therefore are absorbers.
Electron beam deposition (left) produces porous, low-density columnar
coatings subject to water absorption. Ion beam sputtering deposition (right) yields
fully densified coatings with virtually no water absorption.
Damage testing
Unfortunately, despite the best efforts of optics manufacturers
to control their fabrication, cleaning and coating operations so as to maximize
damage resistance, there are still many contributing factors that occur intermittently
and that aren’t well controlled or understood. This means that statistical
quality control methods that rely on relatively small sample sizes can’t be
effectively applied. Thus, extensive testing on virtually every coating run often
must be performed to determine damage characteristics. However, laser damage testing
is time-consuming and destructive, so there is a clear trade-off between establishing
confidence and production cost.
It is important to understand the exact protocols and methods
that a vendor uses to damage test its optics. The number of parts tested, the number
of individual sites tested per part, and the type of part (witness piece or actual
production unit) should all be ascertained. The specifics of the test setup (laser
type, wavelength, pulse energy), the controls used (pulse energy and beam profile
measurements), and compliance with ISO testing standards should all be discussed.
It can also be advantageous to work with a vendor who has in-house damage testing,
because this generally avoids the detrimental effects introduced by product shipping
and handling, and the environment at the test facility; it also enables more cost-effective
testing of a larger number of parts.
As long as manufacturers and users of lasers continue to strive
for ever-higher powers and ever-smaller footprints, laser damage will be with us.
However, working with suppliers to understand the specific steps they take to control
and minimize the sources of damage will give buyers more realistic expectations
and better outcomes.
Meet the author
Trey Turner is vice president of technology at REO Inc. in Boulder,
Colo.; e-mail:
treyt@reoinc.com.
LATEST NEWS