Kanishka Tankala, Jaroslaw Abramczyk, Douglas Guertin and Nils Jacobson, NUFERN
Fiber lasers for industrial, military, scientific and medical applications
are now ubiquitous. Active and passive fibers are used in fiber lasers as gain media,
for numerous fiber-based components and for beam delivery. To date, much of the
discussion on reliability of fiber lasers has been focused on diode and fiber reliability.
In the regime of fiber reliability, attention primarily has been given to photo-darkening
of the double-clad fibers used as gain media. Low-index polymer coatings are key
in guiding the pump light, and their mechanical and environmental reliability has
not received due attention.
Because fiber manufacturers have not universally provided well-engineered
polymer coatings for mechanical and optical reliability, laser manufacturers have
either lived with fibers of questionable reliability or resorted to such methods
as potting the double-clad fibers to create a barrier to moisture ingression and
mechanical damage. And because moisture diffusion is relatively rapid in polymers,
potting the fibers in polymers is expected to improve reliability only marginally.
A more lasting solution is to engineer the low-index polymer to withstand the deleterious
effects of the high temperature and humidity encountered during storage and operation.
A synthetic confocal microscope image of an ytterbium-doped
laser fiber from Nufern. The octagon shape of the glass is shown, as are
the dual polymer coatings carefully stripped back. Courtesy of
A typical double-clad fiber includes a core that carries the signal
light; a first cladding surrounding the core, which carries the pump light; and
a second cladding that helps contain the pump light in the first cladding. Although
the second cladding can be a fluorosilicate glass, the index of such glass can
barely provide numerical apertures (NAs) of 0.30, significantly below the NAs of
≥0.46 typically needed for double-clad fibers. Low-index polymers with refractive
indices of ≥1.37 provide desired NAs of ≥0.46.
Passive double-clad fibers are used to transport pump and signal
light to and from the gain medium (the active fiber) and are used in components
including gratings, pump combiners, taps, filters, isolators and acousto-optic modulators.
In an active fiber, the pump light is transported in a noncircular first cladding
until it is absorbed by the lanthanide dopants in the core. The role of the low-index
polymer is to reliably contain the pump light in the first cladding over the life
of the laser. Low-index polymer-coated fibers not only should be robust to mechanical
handling, but also should reliably perform their function over the temperature and
humidity conditions experienced during storage and operation.
Unlike fibers for telecommunications, where the coating performs
the sole function of providing mechanical protection, polymer coatings used for
double-clad fibers perform both mechanical and optical functions. A dual acrylate
coating system – in which the low-index polymer coating forming the first
coating contacting the glass is followed by a rugged secondary coating to protect
the relatively delicate low-index coating – is essential for double-clad fibers.
The robust secondary coating mechanically protects the low-index coating from nicks
and scratches that can cause light to leak from the fiber, resulting in localized
hot spots or catastrophic burns. Dual-acrylate-coated double-clad fibers can exceed
the stringent mechanical reliability standards set forth by the Telcordia GR-20
standard. Double-clad fibers with a dual acrylate coating can exhibit near-theoretical
strength of silica. Measured typical median failure stress of 715 kpsi and a 15
percent failure stress of 708 kpsi significantly exceed the GR-20 standards of ≥550
kpsi and ≥455 kpsi, respectively. In addition, a stress corrosion parameter of
20 was measured, comfortably meeting the GR-20 requirement of ≥18.
Cladding attenuation changes in a glass fiber coated with low-index polymer. Courtesy of Nufern.
The GR-20 standard provides guidelines for mechanical performance
before and after exposure of fibers to 85 °C and 85 percent relative humidity
for 720 hours (30 days) to ensure mechanical reliability. Although mechanical performance
of polymer-coated glass fibers has been studied extensively in the telecommunications
industry, little if any work has been carried out on the reliability in optical
performance of low-index polymer-coated fibers for the fiber laser industry.
Degradation of low-index polymers with exposure to temperature
and humidity can lead to losses by increased absorption or scattering of pump light,
resulting in a degradation of laser output power. Figure 1 shows attenuation changes
of three glass fibers, with various low-index polymer coatings, exposed to 85 °C/85
percent relative humidity. Coatings A and B seem to show a rapid increase in attenuation
after a few hundred hours of exposure, while Coating C, the fiber with specially
engineered coating, shows only a modest increase, even after 1500 hours of exposure.
It should be noted, however, that this 85 °C/85 percent study
does not explicitly reveal the optical reliability of the low-index polymer-coated
double-clad fibers. More specifically, because OH ingression into glass also increases
background attenuation, it is important to distinguish the coating-related attenuation
changes from increases in attenuation of the glass. Figure 2 shows the spectral
attenuation changes for a glass fiber coated with a low-index polymer. The spectral
curve prior to exposure (0 hours) to an 85 °C/85 percent relative humidity
environment provides the baseline spectral features for comparison.
Spectral attenuation changes in a glass fiber coated with
Coating B with exposure to 85 °C/85 percent relative humidity. Courtesy of Nufern.
The baseline spectrum shows that the attenuation in the typical
pump wavelength range before exposure to temperature and humidity is well below
a negligible 0.01 dB/m. Upon exposure to temperature and humidity, both wavelength-dependent
and -independent attenuation changes are observed. The 940- and 1240-nm attenuation
peaks are attributable to –OH overtones in silica glass, and the attenuation
increase below 800 nm is believed to be originating from glass defects resulting
from moisture ingression. A significant wavelength-independent component of attenuation
also is observed and is attributed to light scattered by the low-index polymer upon
exposure to moisture. The 85 °C/85 percent relative humidity test provides
enough time for moisture not only to degrade the low-index polymer, but also to
penetrate the glass cladding, making it difficult to independently evaluate the
coating performance and to benchmark various low-index polymers.
Increase with time in attenuation for fibers soaked in
85 °C hot water. Courtesy of Nufern.
The diffusion of moisture into a polymer is significantly faster
than into glass, so a shorter duration test at a significantly elevated surface
concentration would be better to evaluate the coating performance. And monitoring
the attenuation at a wavelength such as ~1100 nm, which is minimally affected by
the attenuation peaks related to –OH in glass, is better suited for monitoring
coating performance. The increase in attenuation at 1100 nm in an 85 °C hot
water bath for fibers drawn with two different coatings is shown in Figure 3.
It is observed that fibers with Coating B degrade in a matter
of one to three hours, with attenuations reaching 50 to 300 dB/km. In contrast,
fibers drawn with the specially engineered low-index polymer, Coating C, perform
exceedingly well with negligible increase in attenuation, even after 500 hours of
exposure. Clearly, a well-engineered low-index polymer coating, Coating C, can deliver
two to three orders of magnitude better performance, significantly alleviating reliability
With the increasing deployment of fiber lasers, many component
and laser manufacturers are becoming increasingly conscious of the reliability of
their devices. These users of double-clad fibers are demanding proof of reliable
performance, and some are taking steps to do incoming inspection of fibers.
Performance of three commercially available double-clad
fibers subjected to the hot water soak test. Courtesy of Avensys-ITF Labs.
Avensys-ITF Labs, a supplier of high-power combiners, has adopted
the recommended hot water soak test to qualify fibers from various suppliers. It
performed a 10-hour water soak test to qualify fibers from three suppliers and contributed
Figure 4 for this article to report the usefulness of the test and to encourage
users of double-clad fibers to adopt this proposed standard to distinguish performance
of commercially available double-clad fibers. The company also provided data (Figure
5) to demonstrate that the increased attenuation upon moisture ingression manifests
itself as a severe restriction in the numerical aperture of the fiber. Components
and lasers made with fibers that have poor reliability will experience output power
degradation because of increased insertion losses and, in extreme cases, could fail
The availability of double-clad fibers with well-engineered low-index
polymer coatings has significantly alleviated an important reliability concern for
fiber lasers. Fibers with two to three orders of magnitude better optical reliability
have been engineered and are commercially available. However, fiber manufacturers
are yet unable to use accelerated tests to predict lifetimes in environmental conditions
specified for operation and storage of fiber lasers.
Numerical aperture of the commercially available fibers
after exposure to hot water. Courtesy of Avensys-ITF Labs.
The task is further complicated by the fact that many laser manufacturers
cannot determine the temperature of active fibers and components made with passive
fibers under specified environmental conditions. Lifetime predictions under specific
environmental conditions and laser designs remain a subject for future work.
Meet the authors
Kanishka Tankala is vice president of operations at Nufern; e-mail:
firstname.lastname@example.org. Jaroslaw Abramczyk is technical manager; e-mail: email@example.com.
Douglas Guertin is senior engineer; e-mail: firstname.lastname@example.org. Nils Jacobson is
director of process engineering; e-mail: email@example.com.