Optical Fibers for Industrial Laser Applications
This is a continuation from the previous tutorial - Sapphire optical fibers
1. FIBER LASERS AND AMPLIFIERS: AN INTRODUCTION
Both optical amplifiers and lasers are based on the process of stimulated emission; a concept first proposed by Albert Einstein in 1916 but not demonstrated until 1954 when Charles Townes used stimulated emission to produce microwave oscillation in the ‘‘maser’’ (microwave amplification by stimulation emission of radiation). It was another 6 years later before Theodore Maiman demonstrated the first laser (light amplification by stimulated emission of radiation).
In these devices, an ‘‘optically active’’ species has one of its electrons ‘‘excited’’ into a higher energy level, a passing photon of energy equal to the difference between the electron’s energy and a lower energy state ‘‘stimulates’’ the electron to fall into the lower energy state and thereby emit a second photon. This photon will necessarily have the same energy, optical frequency, and phase as the original photon.
The waves associated with these two photons constructively interfere and the result is a more intense ‘‘amplified’’ optical beam. If the electrons are reexcited into the higher energy level and feedback of the amplified signal is provided via a resonance cavity, then the ‘‘amplifier’’ may become a ‘‘laser.’’
The lanthanide-doped glass fiber laser was invented in the mid 1960s, making it almost as old as the laser itself. However, the complications inherent to their early design have until now restricted their real-world applications.
Moreover, fiber lasers remained for many years significantly inferior to their \(\text{Nd/YAG}\) and gas laser alternative technologies, rendering them little more than a scientific curiosity with but a few minor niche applications.
2. CLADDING PUMPED FIBERS
Early fiber lasers were side-pumped with a flashlamp, but in 1974, Julian Stone and Charles Burrus took the technology a significant step forward when they demonstrated a neodymium-doped multimode fiber laser that was end-pumped with a laser diode.
However, at that time the only available technique for achieving an acceptable optical quality of the laser output was to employ a fiber with a geometrically small core (on the order of a few microns). The need to couple excitation energy directly into this small core meant that the total achievable output power of these devices was limited to the milliWatt range.
With the advent of cladding pump fiber designs in 1988, the limitation to power scaling fiber devices became the availability of high brightness pump radiation rather than the fiber itself.
This trend continued using the available fiber technology, culminating in 1999 with the demonstration of the world’s first single-mode fiber laser exhibiting a continuous-wave \(\text{(CW)}\) output power in excess of 100 \(\text{W}\).
Traditional optical fiber has a core refractive index raised respective to the surrounding cladding material. The coating has a significantly higher refractive index than either the core or the clad and from an optical perspective is designed to strip out higher order cladding modes that might otherwise re-couple with the core modes.
Double-clad fibers \(\text{(DCFs)}\) differ from traditional ‘‘single-clad’’ optical fiber in the fundamental design of a secondary external wave-guiding structure surrounding the inner core waveguide.
For some applications, it is potentially advantageous to be able to splice the fiber directly with the all-glass pump delivery fiber, and in such cases, a triple-clad fiber incorporating a third all-glass 0.23-numerical aperture \(\text{(NA)}\) cladding between the inner cladding and the polymer coating may be used. The differences in design are shown in Fig. 1.
By negating the requirement for excitation energy to be coupled directly into the relatively small single-mode core, \(\text{DCF}\) makes it possible to employ low-cost, large-area (multimode), high-power semiconductor pump sources. The fundamental concept of the \(\text{DCF}\) structure is that low-brightness, high-power multimode diodes yielding tens and even hundreds of Watts can be used to provide pump power for lanthanide-doped fibers that will convert that energy into highbrightness, high-power, potentially single-mode output.
Fibers for high-power laser and amplifier applications require large claddings with high \(\text{NAs}\) for efficiently coupling pump energy. Such fibers are typically available with cladding diameters up to around 1 mm but more commonly around 400 \(\mu m\).
A fluorinated polymer optical cladding typically provides an \(\text{NA}\) of around 0.46 and that is in turn often surrounded by a more standard telecommunications type of jacket (for abrasion resistance).
The choice of cladding diameter is

dictated primarily by the brightness of available pump diodes and the total power being coupled.
The geometry of the inner cladding is typically shaped to prevent the propagation of skew rays that might otherwise pass down the fiber length without traveling through and being absorbed by the doped core region.
3. LARGE-MODE-AREA YTTERBIUM-DOPED FIBERS: THE POWER REVOLUTION
For certain applications, such as ranging and free-space communications, operating in the ‘‘eye-safe’’ 1.5–2.0 \(\mu m\) range is preferred. Furthermore, there are a number of sensing and medical applications that require other specific wavelengths.
For such applications, it becomes necessary to employ various

optically active lanthanide ions, such as neodymium, thulium, or co-doped erbium/ytterbium.
However, for non–wavelength-specific applications requiring only extremely high-output powers, a number of unique advantages have made ytterbium the dopant of choice.
More specifically, ytterbium-doped fibers offer high output powers tunable over a broad range of wavelengths, from around 975 to 1120 nm (typically ~1060 nm).
Ytterbium also has a relatively small quantum defect—that is to say, because the pump wavelength (typically 915–975 nm) is close to the lasing wavelength, very little energy is lost to heating.
Furthermore, unlike other lanthanide ions, ytterbium has only a single excited state and thereby is not subject to complications arising from excited state absorption \(\text{(ESA)}\) and is relatively immune to self-quenching processes.
Consequently, high concentrations of ytterbium ions

can be incorporated while maintaining excellent conversion efficiencies (typically >75%).
For this reason, the industry has focused on the development of ytterbium- doped fibers and the following discussion deals primarily with these fiber designs.
It should be noted, however, that neodymium/ytterbium–co-doped fibers have demonstrated power scaling advantages by virtue of the fact that they increase the options for wavelength multiplexing the pump diodes (the neodymium has a peak absorption ~ 810 nm and gain peak ~1060 nm).
One of the most significant keys to ensuring broad marketability of the fiber laser is to develop a technique for producing ever increasing output powers without sacrificing beam quality.
Naturally it is possible to ensure diffractionlimited beam quality from a single-mode core in a DCF geometry.
Unfortunately, such a design also limits the total achievable output power and in pulsed laser devices the average power, peak power, and pulse energy. These limitations are the result of low energy storage (for pulsed applications, the energy storage capacity is determined by a combination of the number of active species present and the maximum achievable population inversion, which is in turn determined by the likelihood of amplified spontaneous emission \(\text{[ASE]}\)) and the effects of parasitic nonlinear processes.
More specifically conventional small-core, high \(\text{NA}\) fiber designs limit the maximum achievable output power because of their fundamental susceptibility to optical nonlinearities, including stimulated Raman scattering \(\text{(SRS)}\), stimulated Brillouin scattering \(\text{(SBS)}\), and self-phase modulation.
To overcome the limitations imposed by these parasitic nonlinear processes, it has been necessary to develop fibers with high lanthanide concentrations in relatively large-core, low-\(\text{NA}\) fibers.
By increasing the core diameter of a fiber and reducing the core \(\text{NA}\), it is possible to maintain single-mode operation while both reducing the fraction of spontaneous emission captured by the core and decreasing the power density in the fiber, thereby increasing the threshold power for the nonlinear processes. Furthermore, the total number of active ions present and so the energy storage capacity increases as the square of the core diameter (for a given glass dopant concentration and cladding diameter).
Consequently, it is possible to reduce the length of the fiber device thereby further increasing the threshold for the nonlinear processes.
Of course, there is an upper limit to the core diameter beyond which singlemode operation is not guaranteed.
More specifically for a step-index fiber, it is known that single-mode operation requires that the \(V\)-value remains below 2.405, where \(V\) is proportional to the core diameter \(\text{(d}_{core})\) and \(\text{NA}\) \(\text{(NA}_{core})\) and inversely proportional to the wavelength of operation \((\lambda)\):
\[\tag{1}V=\frac{\pi d_{core}\text{NA}_{core}}{\pi}\]
At very low \(\text{NAs}\) (approximately \(<\)0.06), fibers begin to exhibit extremely high bend sensitivity.
This imposes a practical lower limit on \(\text{NA}\) and, hence, an upper limit on core diameter. Fortunately, however, there are a number of techniques for the suppression of higher order lasing modes that allow us to use even larger core diameters, wherein essentially multimoded fibers can be made to operate with a diffraction-limited beam quality.
These techniques include suitably manipulating the fiber index and dopant profiles; using special cavity configurations; tapering the fiber ends; adjusting the seed launch conditions; and coiling the fiber to induce substantial bend loss for all transverse modes other than the fundamental.
Perhaps the simplest and least expensive of these is the coiling technique, which does not require careful matching of the seed mode and does not rely on complex fiber designs.
It is only necessary to choose the radius of curvature (based on core diameter and \(\text{NA})\) that will discriminate against high-order modes. This technique exploits the fact that the fundamental mode is the least sensitive to bend loss and that the attenuation due to bend loss is exponentially dependant on the bend radius.
For example, Fig. 4 shows the bend loss as a function of bend radius for a 0.06 \(\text{NA}\), 30-\(\mu m\) core diameter fiber.
Such a fiber in a linear configuration can support around five modes, but with the appropriate choice of bend radius (say, ~50 mm), the \(\text{LP11}\) experiences around 50 dB/m of attenuation (and higher order modes are even more severely attenuated) while the \(\text{LP01}\) mode experiences only around 0.01 dB/m.
It is important to note that this technique does not involve the stripping of power from higher order modes, but the suppression of those modes along the entire fiber length.
As such, power is not attenuated and the efficiency of the laser device is not markedly reduced.
In Fig. 5, we show the measured near-field spatial profile of an ytterbium-doped fiber amplifier with a core diameter of 25 \(\mu m\) and an \(\text{NA}\) of 0.1 when


seeded with a \(\text{CW}\) laser at 1064 nm. The profile on the left shows the multimoded (~ 27 guided modes) output of the uncoiled fiber and on the right the diffractionlimited (measured \(M^2\) value of 1.08 \(\pm\) 0.03) output of the coiled fiber.
These so-called large-mode-area \(\text{(LMA)}\) fibers are directly responsible for the explosion in demonstrated diffraction-limited beam-quality output powers, now exceeding the kiloWatt level from a single fiber (Fig. 6).
With the advent of this new class of fibers, the power limitations were once again placed on the pump source rather than the fiber.
The advantages of lanthanide-doped \(\text{LMA}\) fibers are realized by understanding the limiting mechanisms of output power for a typical laser or amplifier.
One such mechanism is \(\text{ASE}\), which extracts energy from the fiber in an incoherent

manner. As described earlier, \(\text{LMA}\) fibers have cores with low \(\text{NAs}\), typically smaller than single-mode telecom fibers. This reduction in core \(\text{NA}\) reduces the amount of fluorescence captured by the core and, thus, the reduction of amplification of that fluorescence.
A second mechanism, nonlinear in nature, is \(\text{SBS}\), which results from an acoustic wave formed from the superposition of the propagating light wave and the counter-propagating stokes wave generated from the index modulation in the glass created by the propagating wave.
The threshold power at which \(\text{SBS}\) occurs in single-mode fibers is given by Eq. (2):
\[\tag{2}\boldsymbol{P}_{Th}=21\cdot\frac{A_{eff}}{L_{eff}\cdot\text{g}_B},\]
where \(A_{eff}\) is the effective area of the fiber, \(L_{eff}\) is the effective length of the fiber given by \(L_{eff}=[1-\text{exp}(-aL)]/a\), and \(\text{g}_B\) is the Brillouin gain coefficient.
The larger core sizes (i.e., effective areas) of \(\text{LMA}\) fiber raises the \(\text{SBS}\) threshold, compared to single-mode fibers (by a few orders of magnitude) and enhances the power handling capability of the laser.
For example, when pumped at 915 nm a single-mode ytterbium-doped fiber with a 0.15 \(\text{NA}\), a core diameter of 5 \(\mu m\), and an ytterbium ion concentration of around 1 \(\text{wt}\)% has an \(\text{SBS}\) threshold of around 40 \(\text{W}\) at 3-\(\text{kHz}\) line width, while a 20-\(\mu m\) core 0.06-\(\text{NA}\) fiber has a threshold of around 340 \(\text{W}\) and a 30-\(\mu m\) core 0.06-\(\text{NA}\) fiber has a threshold of around 680 \(\text{W}\).
In practice, the Brillouin pump line-width can range from a few kiloHertz to several megaHertz, so the actual thresholds may be significantly higher depending on the system configuration.
By pumping at 976 \(\text{nm}\) or using more highly doped fiber, even higher \(\text{SBS}\) thresholds may be achieved.
4. POLARIZATION-MAINTAINING LMA DCF
It is not feasible to indefinitely increase the output power capability of an \(\text{LMA}\) \(\text{DCF}\) through scaling of the core diameter. Ultimately there will be some upper limit, above which output beam quality will begin to degrade.
To help overcome this hurdle, research is also underway to further refine the design of

\(\text{LMA}\) \(\text{DCFs}\), through optimization of the glass composition and wave-guiding structure. These include techniques for reducing the peak power density of light propagating in the core, via careful manipulation of the core refractive index profile.
Large-flattened-mode \(\text{(LFM)}\) optical fibers have been designed and fabricated to further increase the threshold for nonlinear processes in \(\text{LMA}\) fiber by homogenizing the power-intensity profile across the core region (Fig. 8).
At very high peak powers, optical damage of the fiber and selffocusing are limitations to power scaling. Beam-expanding endcaps reduce surface damage effects and the increased mode-field area \(\text{(MFA)}\) achieved through \(\text{LFM}\) fiber designs have enabled greater than 1.5-\(\text{MW}\) peak powers (at ~1-\(\text{ns}\) pulse duration) to be achieved.
Despite the very large \(\text{MFA}\) from these fibers, good beam quality (Table 1) has been demonstrated in practical systems. Nevertheless, the effectiveness of such techniques is somewhat limited and alternative techniques are required for significant power-scaling requirements.
\(\text{CW}\) output powers exceeding 1 \(\text{kW}\) have already been demonstrated in multiplexed fiber devices with poor beam quality and near diffraction-limited output powers exceeding 1 \(\text{kW}\) have also been demonstrated.
However, with the growing need for output powers of several kiloWatts for industrial cutting and welding applications and greater than 100 \(\text{kW}\) \(\text{(CW)}\) for military and aerospace applications, the current goal of a number of research groups is to achieve diffraction-limited kiloWatt powers from a single fiber and then to combine the outputs of several such devices.
A number of such power-scaling techniques have been demonstrated including coherent beam combining, spectral beam combining, and polarization beam combining.
For these extremely high-power applications, operation under stable linear polarization is becoming a requirement. Furthermore there are a number of other applications requiring polarization-maintaining \(\text{(PM)}\) output including coherent

Table 1. Power amplifier performance of a large flattened mode (LFM) fiber with around 1.5-MW peak power

optical communications, nonlinear frequency conversion, pumping optical parametric devices, and all manner of mode-locked, Q-switched, and narrow line-width fiber lasers. Consequently, there has been an increasing demand for \(\text{PM}\) \(\text{DCFs}\).
In the past, different approaches have been suggested to obtain \(\text{PM}\) operation using non-\(\text{PM}\) fibers. Such approaches, however, have their limitations and the preferred technology is to use a truly \(\text{PM}\) \(\text{DCF}\).
Although passive \(\text{PM}\) fibers have been commercially available for many years, actively doped \(\text{PM}\) fibers have not been available until recently. In fact an amplifier employing \(\text{Yb}\)-doped \(\text{PM}\) \(\text{DCF}\) was first reported by Kliner et al. in 2001.
This fiber was of bowtie geometry, and though acceptable for proof of concept and research and development, it has substantial limitations in terms of preform manufacturability, uniformity, and scalability.
Furthermore, the nonideal refractive index profile inherent to such doped bowtie fibers (Fig. 9) makes diffraction-limited operation difficult to achieve.
Figure 10a schematically demonstrates the steps involved in making a bowtie type of \(\text{PM}\) fiber. A high-quality synthetic quartz tube is used as a substrate and several layers of borosilicate glass are first deposited on the inner wall of the rotating substrate.
Next the substrate rotation is stopped, and using a specialized ribbon burner, the boron in the glass is volatilized from a selected sector of the deposited layer. The substrate tube is then rotated by 180 degrees and a similar sector is volatilized. Special care has to be taken to ensure that the sectors of glass from which the boron has been volatilized are diametrically opposite to each other and dimensionally equal along the substrate length.
Several layers of glass are further deposited before the doped core is deposited. These layers act as a buffer between the borosilicate stress members and the core and ensure that the evanescent field does not propagate in the stress elements to any significant extent. The actively doped core is typically deposited using a solution doping technology.
The substrate tube with the various layers of deposited glass is then carefully


collapsed into a rod. The collapsed preform is further processed to obtain the desired inner cladding geometry and drawn into a fiber.
\(\text{PANDA}\)-type \(\text{PM}\) \(\text{DCFs}\) are manufactured in two separate stages, as schematically illustrated in Fig.10b. The actively doped preform is fabricated in a separate process and may employ a manufacturing technology more suitable for yielding highly uniform lanthanide and co-dopant distributions.
A high-quality synthetic quartz tube is used to deposit the lanthanide-doped glass. The tube is then collapsed into a rod and further processed so when drawn, the fiber will have the desired core and inner cladding dimensions.
In a separate step, two circular stress elements of desired composition are fabricated. Two holes of the desired dimension are drilled, either side of the core, in the lanthanide-doped preform.
The circular stress members are inserted into the holes and incorporated into the preform. The preform with the stress members is then drawn into a fiber of desired size and geometry.
The bowtie technology offers the advantage of fabricating the stress members and the lanthanide-doped core in a one-step process. In addition, the distance of the stress members from the core can be precisely controlled by the number of buffer layers deposited between the stress layers and the core.
The stress elements can be brought very close to the core, and hence, for a given size and composition of the stress element, a relatively high birefringence may be achieved. However, this technology has several significant disadvantages.
The need to deposit stress elements and a lanthanide-doped core within the same substrate tube limits the ability to independently control the polarization and lasing properties of the fiber. Furthermore, although the stress elements can be brought close to the core, the size of the stress elements that may be deposited is restricted and thereby limits the size of the preform that can be made with a desired birefringence.
In other words, the technology does not lend itself to volume production. Finally, most \(\text{DCFs}\) require a noncircular geometry of the inner cladding, which calls for some processing step such as grinding or thermal processing to obtain a desired geometry.
In the case of a bowtie type of preform, the grinding (or thermal processing) operation has to be conducted with the stress members in place.
\(\text{PM}\) preforms are relatively fragile because of the large amount of stress incorporated in the preform and are, therefore, prone to fracture on exposure to mechanical (or thermal) shock during a grinding (or thermal processing) operation. The bowtie preform technology is, therefore, not preferred for making volume production of \(\text{PM}\) \(\text{DCF}\).
The technology used to make \(\text{PANDA}\)-type \(\text{PM}\) \(\text{DCF}\) not only offers several advantages but addresses the limitations of the bowtie technology.
In this process, both the lanthanide-doped preform and stress member fabrication steps are effectively decoupled, providing independent and highly effective control of the polarization properties and composition of the lanthanide-doped glass. Furthermore, relatively large stress-inducing members may be fabricated, which substantially increases the limit of preform size and makes the process more suitable for preform scale up.
Finally, all processing required to achieve a noncircular geometry may be accomplished before incorporating the stress members and, hence, improving production yields.
The \(\text{PANDA}\)-type \(\text{PM}\) technology is, therefore, amenable to fabricating \(\text{PM}\) \(\text{DCF}\) and is the technology of choice for reproducible and uniform volume production.
The \(\text{PM}\) ability of all \(\text{PM}\) fibers relies on residual stress anisotropy across the core, which in turn arises from differences in thermal expansion coefficient between the stress members, core and cladding.
The composition, location, and geometry of the stress members determine the birefringence in the fiber. In \(\text{PM}\) \(\text{DCFs}\), the core and cladding geometries are very different to standard telecommunications type of \(\text{PM}\) fibers; more specifically in \(\text{LMA}\) \(\text{DCFs}\), the large diameter of the core negatively affects the achievable birefringence.
Before the feasibility of \(\text{PANDA}\)-type \(\text{PM}\)-\(\text{LMA}\) \(\text{DCFs}\) could first be demonstrated, considerable research had to be performed to optimize the compositional and the geometrical design of the stress members, and in 2003, the results of such detailed experimental and theoretical analyses were reported. Figure 11 shows the key dimensional parameters that determine the birefringence that can be obtained in a \(\text{PM}\) \(\text{DCF}\).
These include the size of the stress member \((d_s)\) and the position of the stress member \((d_p)\) relative to the inner cladding diameter \((d_f)\) and the core diameter \((d_c)\). In addition to the geometric factors, the composition of the stress rod determines the birefringence that is achieved in the fiber.
Figure 12 shows the effect of stress rod size and location on the birefringence (and beat length) of the fiber. As can be seen, the birefringence may be increased (or the beat length reduced) by increasing the size of the stress members \((d_s)\) and keeping all other parameters constant.
Similarly, the birefringence may be increased by moving the stress rods closer to the core.
Although it is theoretically possible to use these two geometric parameters to achieve very large values of birefringence, a limiting criterion imposed on \(d_s\) and \(d_p\) is the distance of the stress members from the core.
This limiting distance is


indicated by the distance between the inside edges of the stress members \((d_i)\). If \(d_i\) becomes very small, the probability of overlap between the mode field and the stress members increases, resulting in increased attenuation and bend loss of the laser or amplifier signal wavelength.
To provide a safety margin for avoiding any overlap between the modal power profile in the fiber and the stress members, it is necessary to determine a critical ratio \(d_i/\text{MFD}\) so that losses are minimized.
For small-core single-mode fibers used in low to medium power applications, it is possible to achieve sufficient birefringence using standard stress member compositions and operate well within the limiting ratio.
However, for large-core fibers as needed for high-power applications, achieving sufficient birefringence while operating within the limiting ratio is more challenging. In such cases, a higher coefficient of thermal expansion difference and, hence, higher birefringence can be achieved by adjusting the composition of stress members so they are similar to those used for gyroscope fibers.
Indeed a broad range of ytterbium-doped \(\text{LMA}\) \(\text{DCFs}\), whose characteristics are optimized for various output powers, are now commercially available. An optical image showing the cross-section of such a fiber, with a 20-\(\mu m\) core and 400-\(\mu m\) inner-cladding diameter and a birefringence exceeding \(3.5\times10^{-4}\), is presented in Fig. 2.
5. FIBER LASERS: STATE OF THE ART
\(\text{LMA}\) fibers with core diameters of 20–30 \(\mu m\) and \(\text{NAs}\) of around 0.06 have become the industry standard for high-power laser and amplifier devices because of their ability to deliver good beam quality through preferential modal excitation or coiling induced higher order mode losses.
The addition of \(\text{PANDA}\)-type stress elements to make \(\text{PM}\)-\(\text{LMA}\) fibers has added to the application space for the fiber technology and enabled high-power linearly polarized fiber amplifiers both in the \(\text{CW}\) and pulsed regimens.
The availability of fibers with large claddings (400 \(\mu m)\) and high cladding \(\text{NAs}\) (0.46) in conjunction with high brightness pump sources has featured in many of the high-power results.
More particularly, they have facilitated the amplification of single-frequency sources into the high-power regimen (hundreds of Watts and as such are potential building blocks for coherent beam-combining and fiber array phase-locking experiments.
An indicator of maturity in the \(\text{LMA}\) fiber technology is the availability of standard support components with \(\text{LMA}\)-compatible fiber pigtails, including the multimode pump combiners, which also serve as signal multiplexers.
These components are available with input fibers compatible with industry-standard pigtail fibers on commercial high-power diodes. For example, the \((6+1)\)-to-1 design consisting of \(\text{LMA}\)-compatible 20/400 \(\text{DCF}\) on the output of the combiner with six 200/220 0.22-\(\text{NA}\) pump delivery fibers on the input side are commercially available.
Furthermore high-brightness, fiber-coupled pump diodes compatible with pump combiners are now commercially available with industrial- grade reliability. These advancements in high-brightness pumps and high-power pump combiners have enabled the high-power, monolithic design shown in Fig. 13.
Experimental results for the system are presented in Fig. 14 and demonstrate the applicability of these high-power \(\text{LMA}\) monolithic amplifiers to output powers greater than 200 \(\text{W}\) \(\text{CW}\).
Although the power level is well below that demonstrated with broad line-width fiber lasers and amplifiers, these \(\text{LMA}\) devices are applicable to amplifying single-frequency input signals with

coherence lengths suitable for further beam combining into the multi-kiloWatt regimen.
\(\text{PM}\) versions of these \(\text{LMA}\) fibers have also been demonstrated to exhibit excellent slope efficiencies and operate at high powers, greater than 400 \(\text{W}\) pump power limited.
Indeed, Nufern scientists have combined the \(\text{PM}\)-\(\text{LMA}\) fiber concept with an optimized coil form to deliver a unique linearly polarized fiber laser, as shown schematically in Fig. 15. Alternative methods for delivering polarized fiber lasers inevitably include external or free space components, or at the very least extra polarizing components within the cavity.
By simply optimizing the fiber and coil combination, it is possible to make a high-power polarized fiber laser by taking advantage of the difference in bend-induced attenuation for each of the two polarization states.
The excellent polarization/extinction ratio, greater than 95%, was obtained with diffraction-limited beam quality, and a grating stabilized line width of around 0.1 nm. Importantly, the output power


was limited only by the available pump power, and in fact it is anticipated that the maximum \(\text{CW}\) output power for this fiber design will be around 2 \(\text{kW}\).
\(\text{LMA}\) fibers have also had a dramatic impact on pulsed fiber laser technology where pulse energies are now approaching 100 \(\text{mJ}\) with multimode beam quality and around 4 \(\text{mJ}\) single mode.
High peak powers, in combination with the short pulse durations (nanoseconds) and diffraction-limited beam quality, are desired in many applications including marking, micromachining, and drilling anything from fuel injectors through to turbine blades.
Military applications involve target designation and laser radar, which is also a growing commercial market with applications such as vehicle guidance, robotics/vision systems, metrological, and surveying.
The combination of high average power (over hundreds of Watts), excellent beam quality, and a polarized output has caught the attention of the \(\text{DPSSL}\) community.
Such performance specifications are extremely difficult to achieve in either \(\text{CW}\) or pulsed \(\text{YAG}\)/Vanadate lasers where fundamental problems such as thermal lensing push resonator design to the limits.
Furthermore, the flexibility of monolithic components make concatenating amplifier stages relatively easy in fiber devices, opening the potential for highly flexible devices generating pulse durations from sub-nanosecond to \(\text{CW}\) and pulse energies in the milliJoule range.
6. LARGE-MODE-AREA EYE-SAFE FIBERS
As discussed earlier, new developments in \(\text{LMA}\) fibers have led to demonstrations of kiloWatt-level outputs in \(\text{CW}\) lasers and megaWatt-level peak powers in pulsed amplifiers (with sub-nanosecond pulses).
However, development of \(\text{LMA}\) fibers has largely been restricted to ytterbium-based fibers for use at around 1.0 \(\mu m\), because of the relative ease of manufacturing \(\text{LMA}\) ytterbium-doped fibers and high optical efficiencies (~80%), associated with this laser system.
The ability to achieve relatively high concentrations of ytterbium with relatively low levels of matrix-modifying co-dopants lends itself to fabricating low-\(\text{NA}\), largecore fibers. The low-\(\text{NA}\) core supports only a few modes and the higher order modes can be easily discriminated against by preferential seeding or bending to achieve diffraction-limited operation.
In spite of the numerous advantages, a significant drawback of the ytterbium-based system is the relatively high sensitivity of the human eye to wavelengths in the 1.0-\(\mu m\) region.
The retinal absorption of radiation in the 1.5- and 2-\(\mu m\) wavelength regions is substantially lower than that at 1.0 \(\mu m\). High-power lasers and amplifiers operating in these wavelength bands are, therefore, of interest in both military and commercial applications such as free-space and satellite optical communications and \(\text{LIDAR}\).
In military applications, these wavelengths are also of interest because they minimize collateral damage. Similarly, eye-safe lasers are welcome in commercial application because they greatly reduce the challenges associated with laser safety.
In addition to eye safety, the 2.0-\(\mu m\) lasers may also find applications as pump source for holmium-doped lasers or nonlinear conversion to longer wavelengths.
Although there has been significant interest in eye-safe lasers, progress in developing high-power \(\text{CW}\) and pulsed lasers at these wavelengths has been limited by the availability of \(\text{LMA}\) erbium/ytterbium– and thulium-doped fibers.
This limitation is itself due to the difficulties in manufacturing optically efficient low-\(\text{NA}\) optical fibers containing these lanthanide dopants.
Despite the lack of LMA fibers, important work has been conducted in developing high-power erbium/ytterbium lasers using multimode fibers. Koroshetz et al. demonstrated a 40 \(\text{W}\), 10-\(\text{Gbps}\) amplifier for free-space communication and Shen et al. reported 188 \(\text{W}\) of \(\text{CW}\) output with an \(M^2\) of 1.9 using a 30-\(\mu m\), 0.22-\(\text{NA}\) erbium/ytterbium–co-doped fiber.
Furthermore, Yusim et al. have been able to take a 20-\(\mu m\) core erbium/ytterbium fiber, which supports 30 modes, and achieve 100 \(\text{W}\) output with a near diffraction-limited beam by employing single-mode fiber-based components in the cavity and making careful splices between the single-mode fibers and the multimode active fiber.
Although a high-power diffraction-limited output was achieved, such methods are cumbersome and emphasize the need for \(\text{LMA}\) fibers. Similarly, noteworthy work has been conducted in developing highly efficient thulium-doped fibers, by exploiting the two-for-one cross-relaxation process between thulium ions and output powers as high as 85 \(\text{W}\) have been demonstrated albeit with a multimode output.
Unlike ytterbium-doped fibers, the compositional requirements for erbium/ ytterbium–co-doped and thulium-doped fibers make fabrication of \(\text{LMA}\) variants particularly challenging.
More specifically, it is well known that sensitizing erbium-doped fibers with ytterbium enhances pump absorption and hence increases the efficiency at the erbium lasing wavelength.
Erbium has a very narrow absorption peak and, like any lanthanide ion, cannot be incorporated into silica glass at extremely high concentrations without clustering. Sensitization is accomplished by taking advantage of the broad absorption band and the high cross-section of ytterbium as compared with erbium, with the net result that erbium/ ytterbium–co-doped fibers have a very broad absorption band with a peak absorption that is more than two orders of magnitude greater than conventional erbium-doped fibers.
For efficient energy transfer from the ytterbium to erbium ions, the Raman shift of the base glass is increased by doping it with phosphorus. The presence of \(P=O\) bonds increases the phonon energy of the glass host, and the Raman spectrum has a peak at 1330 \(\text{cm}^{-1}\), as compared to 1190 \(\text{cm}^{-1}\) for pure silica.
This helps in rapid depopulation of the erbium \(I_{11/2}\) energy level, limiting the back-transfer of energy from erbium to ytterbium ions (Fig.16).
Thus, efficient erbium/ytterbium fibers contain phosphorus in the core, which also aids in minimizing the clustering of the lanthanide ions.
A substantially high level of phosphorus is required to achieve both of the

aforementioned goals.
However, phosphorus also increases the refractive index of the base glass, resulting in relatively high core \(\text{NAs}\) of around 0.17–0.20 or higher.
Thus, the difficulty in producing an \(\text{LMA}\) fiber, which requires a low \(\text{NA}\), becomes apparent.
Thulium-doped fibers have a number of potential pump bands (Fig. 22.17). However, pumping at 793 nm is preferred because of the availability of highpower semiconductor diodes in the 800-nm spectral region. In conventional thulium-doped fibers, the maximum conversion efficiency of a 793-nm pumped laser is about 40%.
However, compositionally engineered thulium-doped silica fibers can achieve substantially higher efficiencies (approaching 80%). Important energy transfer processes relevant to the performance of thulium-doped silica fibers have been identified.
Figure 16 shows the relevant cross-relaxation mechanism observed in thulium-doped silica fibers. Two cross-relaxation processes, namely \(^3H_4\), \(^3H_6\rightarrow^3F_4\) \(^3F_4\) and \(^3H_4\), \(^3H_6\rightarrow^3H_5\), \(^3F_4\), have been identified with the \(^3H_4\), \(^3H_6\rightarrow^3F_4\), \(^3F_4\) being particularly efficient because of the large degree of spectral overlap between the \(^3H_4\), \(\rightarrow^3F_4\) emission and the \(^3H_6\rightarrow^3F_4\) absorption.
This cross-relaxation process results in the generation of two signal photons for every pump (793-nm) photon and may be promoted by using high thulium concentrations.
However, energy transfer up-conversion processes, namely \(^3F_4\), \(^3F_4\rightarrow^3H_5\), \(^3H_6\) and \(^3F_4\), \(^3F_4\rightarrow^3H_4\), \(^3H_6\), have to be kept in check to prevent quenching of the \(^3F_4\) energy multiple.
This can be minimized by using very high alumina:thulium concentrations and thereby

preventing clustering of the thulium ions. Therefore, both high thulium and high alumina concentrations are required to achieve greater than 100% quantum efficiencies in thulium-doped fibers.
In fact, careful compositional engineering can yield power conversion efficiencies similar to that of ytterbium-doped fibers, making such fibers extremely desirable for eye-safe laser system.
However, these high dopant concentrations substantially increase the refractive index of the core as compared with the pure silica substrate, and typical \(\text{NAs}\) are in the range 0.18–0.24.
Hence, as in the case of erbium/ytterbium–co-doped fibers, the compositional requirements for an efficient thulium-doped fiber also limits the ability to manufacture low \(\text{NA}\) \(\text{LMA}\) fibers.
In a typical step-index fiber, the \(\text{NA}\) of the core is defined by the index of the core and that of the cladding surrounding it. However, if an appropriate pedestal is designed, the core may have an effective index as defined by the index of the core and that of the pedestal, as shown in Fig. 18.
The pedestal index may be chosen so as to define an effective core \(\text{NA}\) of less than 0.1. Similarly the pedestal size may be chosen so that further increasing the pedestal diameter has no significant effect on the core modes, particularly the fundamental mode.
At this diameter, the pedestal behaves as a ‘‘true’’ cladding to the core, rather than an extended core feature. If the pedestal is made any smaller, the mode-field diameter of the core fundamental mode starts decreasing, and core light may be coupled to the pedestal modes.
At the same time, the pedestal should not be made much larger than required, because of increased manufacturing cost, as well as the possibility of trapping pump light in the helical modes in the pedestal.
The first demonstration of highly efficient large core (25 \(\mu m)\) \(\text{LMA}\) ytterbium/ erbium–co-doped and thulium-doped \(\text{DCFs}\) was recently reported.
In both cases, the core \(\text{NAs}\) were chosen to be around 0.1 and the fibers were drawn to 300 and 250 \(\mu m\) cladding diameters, respectively, with a cladding \(\text{NA}\) of 0.46.
Figure 19 shows the absorption and about 30% slope efficiency for the erbium/ytterbium 25/300 pedestal fiber.
The key benefit of the pedestal fiber design is clearly the lower number of modes supported by the doped core of the fiber.
Consequently, it should be easier to excite and maintain the fundamental mode in this fiber than in a fiber with a similar diameter high \(\text{NA}\) (~0.17) core.
The \(\text{V}\)-number of a 25-\(\mu m\) core fiber at an \(\text{NA}\) of 0.17 is 8.616, while at an \(\text{NA}\) of 0.10, it is 5.067.
Figure 20 shows that incorporation of the pedestal is expected to reduce the core modes from 11 to 4, making it possible to achieve near diffraction-limited beam quality.
Importantly, both exciting the fundamental mode (at the expense of higher order modes) and maintaining the power within this mode through the appropriate fiber length should be significantly easier in the few-moded fiber as compared with the highly multimode fiber, because of reduced mode coupling.
Experimentally, good beam quality from the pedestal fiber has been verified in a seeded amplifier configuration, in which care was taken to excite the



fundamental mode of the doped core using standard mode matching techniques. In fact it was found that beam quality was easier to maintain in this fiber than in a fiber with a high 0.17-\(\text{NA}\) core, even though the core was substantially smaller (18 \(\mu m)\).
Single-mode beam profiles and an example \(M^2\) measurement are shown in Fig.21 for a fiber length appropriate for making efficient amplifiers. Although good beam quality has been demonstrated in laboratory


based experiments with high \(\text{NA}\) large core fibers, developing components and designing systems to guarantee a beam quality specification would be particularly challenging.
It is believed that the adoption of a pedestal fiber design will make the challenges associated with delivering good quality beam profiles at eye-safe wavelengths manageable.
Compositional development of thulium-doped fiber has also been performed to optimize the thulium and alumina concentrations required to promote the two-for-one cross-relaxation process, and slope efficiencies as high as 68% (170% quantum efficiency) have been reported [52] (Fig. 22).
An index profile of the \(\text{LMA}\) thulium-doped fiber \(\text{(LMA}\)-\(\text{TDF}\)-25/250) showing the refractive indices of the core and the pedestal relative to the cladding is also shown in Fig. 22.
The absorption of the fiber at 790 nm was measured to be around 4.5 \(\text{dB/m}\) and the absorption spectrum is shown in Fig. 17. The fiber was demonstrated to have a near–diffraction-limited beam quality and an \(M^2\) of less than 1.3 was reported.