Diffraction-limited 976nm lasers can be used to core-pump ultrafast fiber lasers to overcome nonlinearities with significantly shortened fiber lengths or to pump ultrafast solid-state lasers for much improved beam quality. In order to make Yb fiber lasers operate efficiently as a three-level system at ~976nm, it is critical to use double-clad fibers with large core-to-cladding ratio and additional spectral filters, such as dichroic mirrors in free space or fiber Bragg gratings in all-fiber configurations, to suppress lasing at longer wavelengths. Diffraction-limited 94W at 976nm was achieved in 2008 with an efficiency of ~50% with respect to the launched pump powers at ~915nm using a rod-type PCF and multiple dichroic mirrors. However, the results from flexible fibers with the potential to be used for monolithic fiber lasers are far worse. In this work, the Yb-doped double-clad all-solid photonic bandgap fiber has a core diameter of ~25μm and a cladding diameter of ~125μm. The photonic bandgap was engineered to have its long wavelength band edge just beyond 976nm to suppress lasing at longer wavelengths. We demonstrate a record efficiency of ~54% with regard to the coupled pump power at ~915nm. Pump-limited 38W at ~977nm was achieved with a M2 of ~1.24. ASE at ~1026nm was suppressed by <30dB at all powers. This is possible due to the use of all-solid photonic bandgap fibers which provide both the necessary large core-to-cladding ratio and the additional suppression of the four-level system by strong out-of-band transmission loss.
Thermally induced transverse mode instability (TMI) has been recognized as one of the major limits to average power scaling of single-mode fiber laser. Mitigating the thermal load in single-mode high-power fiber lasers by operating lasing closer to the pump wavelength is one of the effort directions. Here, we demonstrate 220w single–mode output power at 1018nm from an ytterbium-doped all-solid photonic bandgap fiber (ASPBF) pumped at 976nm. The quantum defect is only 4.1%, helping to mitigate the thermal load. The ASPBF fiber has the multiple-cladding-resonant design, leading to better higher-order modes (HOM) suppression in its ~50µm core. The large core/cladding ratio also benefits the 1018nm lasing, providing the higher cladding pump absorption so shorter fiber length is needed with better ASE suppression at longer wavelength. In addition, the use of a phosphosilicate host in this fiber also enhances ytterbium gain at 1018nm, leading to a reduction in the required inversion, further increasing efficiency. In the laser test, one end of fiber is spliced to a high-reflective fiber-Bragg-grating at 1018nm and the other end is right-angle cleaved. ~62% and ~77% lasing efficiency has been achieved around maximum power with respective to the launched and absorbed pump power. The M2 was measured at 130W as 1.06 and 1.17 with respective to the x and y axis.
Transverse mode instability (TMI) has been recognized as a major limit to average power scaling of single-mode fiber laser besides the optical nonlinear effects. One key to mitigate TMI is to suppress the higher-order modes (HOMs) propagation in the optical fiber. By implementing additional cores in the optical fiber cladding, HOMs can be resonantly coupled from the main core to the surrounding cladding cores, leading to better HOMs suppression. Here, we demonstrate an Yb-doped multiple-cladding-resonant all-solid photonic bandgap fiber with a ~60μm diameter core for high power fiber lasers. The fiber has a multiple-cladding-resonant design in order to provide better HOMs suppression. Maximum laser power of 910w is achieved for a direct diode-pumped fiber laser without TMI with a 9m long fiber at 60cm coil diameter, breaking the TMI threshold of 800w that has been observed in large-mode-area PCFs with ~40μm core. This result is limited by fiber end burning due to the un-optimized thermal management. Later experiment demonstrates maximum laser power of 1050w with 90% lasing efficiency versus absorbed pump power in a 8m long fiber coiled at 80cm diameter, limited by the pump source. However, the fiber bending condition needs to be optimized in order to produce a better laser beam quality.
Thermal management is critical for kw-level power lasers, where mode instability driven by quantum defect heating is a major challenge. Tandem pumping using 1018nm fiber lasers are used to enable both high brightness and low quantum defect. It is, however, difficult to realize efficient 1018nm YDFL. The best demonstration to date is limited by the use of both conventional aluminosilicate host and smaller core diameters. In these cases, higher inversion is required due to the aluminosilicate host and higher pump brightness is required due to the smaller core, which results in high signal brightness for the same output power. These factors lead to large pump power to exit fiber, resulting in poor efficiency. Phosphosilicate host, on the other hand, requires much lower inversions to reach the gain threshold at 1018nm. The combination of phosphosilicate host and large-core leakage channel fibers (LCF) is a perfect candidate for efficient 1018nm fiber laser. We report a highly efficient Yb-doped phosphosilicate LCF laser with a quantum defect of 4.1% using a ~50μm-core diameter and ~420μm cladding diameter. The slope efficiency with respect to the launched pump power at 1018nm is 70%. The ASE suppression is <60dB. The large cladding of 420μm demonstrates a combination of high efficiency, ~4% quantum defect and high-power low-brightness diode pumping. We have also studied the limits of operating ytterbium fiber lasers at shorter wavelengths and found the efficiency to fall off at shorter wavelengths due to the much higher inversions required.
Polarizing optical fibers are important components for building compact fiber lasers with linearly polarized laser output. Conventional single-mode optical fibers with birefringence can only preserve the polarization when the incident beam is launched properly. Recent reports demonstrate that the birefringence in photonic bandgap fibers (PBFs) can provide single-polarization operation near the edge of transmission band by shifting the transmission band for the light with orthogonal polarizations. Here, we demonstrate a 50μm core Yb-doped polarizing photonic bandgap fiber (PBF) for single-polarization operation throughout the entire transmission band from 1010nm to 1170nm with a polarization extinction ratio (PER) of >5dB/m, which is >15dB/m near the short wavelength edge of the transmission band. The polarizing effect is due to the differential polarization transmission loss presented in this fiber, which is benefited from the fiber birefringence of 3.2x10-4, obtained by incorporating low-index boron-doped rods on either side of the core. The achievement is based on the fact that light at fast axis has lower effective mode index which is closer to the modes in the photonic cladding and thus to be easily coupled into cladding. A 2.6m long straight fiber was tested in a laser configuration without any polarizers to achieve single polarized laser output with a PER value of 21dB at 1026nm lasing wavelength.
Ytterbium-doped large mode area all-solid photonic bandgap fiber amplifiers were used to demonstrate <400 W of
output power at 1064 nm. In an initial set of experiments, a fiber with a core diameter of ~50 μm, and a calculated
effective area of 1450 μm2 in a straight fiber, was used to generate approximately 600 W. In this case, the input
seed was modulated using a sinusoidal format at a frequency of 400 MHz. The output, however, was multimode as
the fiber design did not allow for single-mode operation at this wavelength. A second fiber was then fabricated to
operate predominantly in single mode at 1064 nm by having the seed closer to the short wavelength edge of the
bandgap. This fiber was used to demonstrate 400 W of single-frequency output with excellent beam quality. As the
signal power exceeded 450 W, there was significant degradation in the beam quality due to the modal instability.
Nevertheless, to the best of our knowledge, the power scaling results obtained in this work far exceed results from
prior state of the art all-solid photonic bandgap fiber lasers.
Fiber lasers are in the process of revolutionizing modern manufacturing. Further power scaling is still much desired to increase throughput and to break new frontiers in science and defense. It has become very clear now that highly single-mode fibers with large effective mode areas are required to overcome both nonlinear effects and mode instability [1-3]. We have been studying all-solid photonic bandgap fibers (AS-PBF), which have open and highly dispersive cladding, making them ideal for higher-order-mode controls in large-mode-area fibers. I will review our recent progress in this area and, especially in ytterbium-doped AS-PBF lasers and amplifiers.
Power scaling of fiber lasers is highly desirable in many applications but is mainly limited by nonlinear effects. Large-mode-area fibers have been used to mitigate this limit, such as the leakage channel fiber (LCF). The mode intensity profile in these fibers typically exhibits Gaussian-like structure with much reduced effective mode-area compared to the physical fiber core area. Thus, a flat-top mode with a uniform intensity distribution is more suitable for larger effective mode-area without having to increase core size. In this work, we demonstrate the first flat-top mode generated in a 50 μm-core Yb-doped LCF fiber. The mode flattening from Gaussian beam to a flat-top one is achieved by using a 30 μm uniform Yb-doped area in the core center with a refractive index very slightly below that of the background silica glass by 2×10-4. The resulting flat-top mode has a significantly increased effective mode area of ~1880 um2, which is ~50% larger than that of a conventional uniform core and ~6 times the effective mode area of the flat-top mode record demonstrated previously. A 6m-long fiber is also tested in a laser configuration with a slope efficiency of ~84% at 1026 nm with respect to the absorbed pump power at 976 nm.
Described herein, for the first time to the best of our knowledge, are results on optical fibers possessing significant compositional gradations along its length due to longitudinal control of the core glass composition. More specifically, MCVD-derived germanosilicate fibers were fabricated that exhibited a gradient of up to about 0.55 weight % GeO2 per meter. These gradients are about 1900 times greater than previously reported fibers possessing longitudinal changes in composition. The refractive index difference is shown to change by about 0.001, representing a numerical aperture change of about 10%, over a fiber length of less than 20 m. The lowest attenuation measured from the present longitudinally-graded fiber (LGF) was 82 dB/km at a wavelength of 1550 nm, though this is shown to result from extrinsic process-induced factors and could be reduced with further optimization. The stimulated Brillouin scattering (SBS) spectrum from the LGF exhibited a 4.4 dB increase in the spectral width, and thus reduction in Brillouin gain, relative to a standard commercial single mode fiber, over a fiber length of only 17 m. The method employed is very straight-forward and provides for a wide variety of longitudinal refractive index and acoustic velocity profiles, as well as core shapes, which could be especially valuable for SBS suppression in high-energy laser systems. Next generation analogs, with longitudinally-graded compositional profiles that are very reasonable to fabricate, are shown computationally to be more effective at suppressing SBS than present alternatives, such as externally-applied temperature or strain gradients.
There are still very strong interests for power scaling in high power fiber lasers for a wide range of applications in medical, industry, defense and science. In many of these lasers, fiber nonlinearities are the main limits to further scaling. Although numerous specific techniques have studied for the suppression of a wide range of nonlinearities, the fundamental solution is to scale mode areas in fibers while maintaining sufficient single mode operation. Here the key problem is that more modes are supported once physical dimensions of waveguides are increased. The key to solve this problem is to look for fiber designs with significant higher order mode suppression. In conventional waveguides, all modes are increasingly guided in the center of the waveguides when waveguide dimensions are increased. It is hard to couple a mode out in order to suppress its propagation, which severely limits their scalability. In an allsolid photonic bandgap fiber, modes are only guided due to anti-resonance of cladding photonic crystal lattice. This provides strongly mode-dependent guidance, leading to very high differential mode losses. In addition, the all-solid nature of the fiber makes it easily spliced to other fibers. In this paper, we will show for the first time that all-solid photonic bandgap fibers with effective mode area of ~920μm2 can be made with excellent higher order mode suppression.
Further power scaling of single frequency fiber lasers is of significant interests for many scientific and defense applications. It is currently limited by stimulated Brillouin scattering (SBS). In recent years, a variety of techniques have been investigated for the suppression of SBS in optical fibers. A notable example is to design transverse acoustic properties of optical fibers in order to minimize optical and acoustic mode overlap. It was pointed out recently that SBS suppression from such transverse acoustic tailoring is limited when considering the existence of acoustic leaky modes. We demonstrate, for the first time, a post-processing technique where hydrogen is diffused in to a fiber core and then locally and permanently bonded to core glass by a subsequent UV exposure. Large local acoustic property can be altered this way for significant SBS suppression. It is also possible to use this technique to implement precisely tailored acoustic properties along a fiber for more optimized SBS suppression in a fiber amplifier. Change in Brillouin Stokes frequency of ~320MHz at 1.064μm has been demonstrated using hydrogen, corresponding to a SBS suppression of ~8dB. Much higher SBS suppression is possible at higher hydrogen concentrations.
There are still very strong interests for power scaling in high power fiber lasers for a wide range of applications in
medical, industry, defense and science. In many of these lasers, fiber nonlinearities are the main limits to further scaling. Although numerous specific techniques have studied for the suppression of a wide range of nonlinearities, the fundamental solution is to scale mode areas in fibers while maintaining sufficient single mode operation. Here the key problem is that more modes are supported once physical dimensions of waveguides are increased. The key to solve this problem is to look for fiber designs with significant higher order mode suppression. In conventional waveguides, all modes are increasingly guided in the center of the waveguides when waveguide dimensions are increased. It is hard to couple a mode out in order to suppress its propagation, which severely limits their scalability. In an all-solid photonic bandgap fiber, modes are guided due to anti-resonance of cladding photonic crystal lattice. This provides strongly modedependent guidance, leading to very high differential mode losses. In addition, the all-solid nature of the fiber makes it easily spliced to other fibers. In this paper, we will show for the first time that all-solid photonic bandgap fibers with effective mode area of ~800m2 can be made with excellent higher order mode suppression.
There are very strong interests for power scaling in high power fiber lasers for a wide range of applications in medical,
industry, defense and science. In many of these lasers, fiber nonlinearities are the main limits to further scaling.
Although numerous specific techniques have studied for the suppression of the wide range of nonlinearities, the
fundamental solution is scaling mode areas in fibers while maintaining sufficient single mode operation. Here the key
problem is that more modes are supported once physical dimensions of waveguides are increased. There are two basic
approaches, lower refractive index contrast to counter the increase of waveguide dimension or/and introduction of
additional losses to suppress higher order modes. Lower index contrast leads to weak waveguides, resulting in fibers no
longer being coil-able. Our research has been focused on designs for significant higher mode suppression. In
conventional waveguides, modes are increasingly guided in the center of the waveguides when waveguide dimensions
are increased. It is hard to couple the modes out to suppress them. This severely limits the scalability of all designs based
conventional fibers. In an all-solid photonic bandgap fiber, modes are guided due to anti-resonance of cladding photonic
crystal lattice. This leads strongly mode-dependent guidance. Our theoretical study has shown that it can have some of
the highest differential mode losses among all designs with equivalent mode areas. Our design and experimental works
have shown the potential of this approach for all-glass fibers with >50μm core which can be coiled for high power
Progress in advanced specialty fibers is the foundation to further breakthroughs in fiber lasers. Recently, we have been
working to advance several areas of developments in specialty fibers and would like to review these efforts here. The
first topic is in the further development of all-glass large core leakage channel fibers (LCF) for robust and practical
solutions for power scaling. The second area is the development of wide band air-core fibers with an innovative square
lattice cladding and the demonstration of a factor of two improvements in bandgap over conventional hexagonal lattice.
These air-core fibers are critical for fiber delivery solution of both CW and pulsed fiber lasers in the future. The last
topic is a new development in design and simulation of SBS gains in optical fibers by incorporating leaky acoustic
modes. These leaky acoustic modes have been mostly overlooked so far. It is essential that they are considered in SBS
simulations in fibers, because they are normal solutions to the acoustic waveguide equations and have similar loss to
guided acoustic modes where the acoustic mode loss is dominated by material loss. This leads to much improved
resolution of SBS gain spectrum in fibers and to new design insights to the limit of SBS suppression based on anti-guide
acoustic waveguide designs.
Air-core photonic bandgap fibers offer many unique properties and are critical to many emerging applications. A notable
property is the high nonlinear threshold which is the key for applications at high peak powers. The strong interaction of
light and air is also essential for a number of emerging applications, especially those based on nonlinear interactions and
spectroscopy. For many of those applications, much wider transmission bandwidths are desired to accommodate a wider
tuning range or a large number of optical wavelengths involved. All demonstrated air-core photonic bandgap fibers so far
have a cladding of hexagonal lattice. The densely packed geometry of hexagonal stacking does not allow large nodes in
the cladding, which are essential for a further increase of photonic bandgaps. On the other hand, a photonic cladding with
a square lattice can potentially provide much larger nodes and consequently wider bandgap. In this work, the potentials
of much wider bandgap with square lattice cladding are theoretically studied.
Two factors are the keys to the rapid commercial development of fiber lasers in the last decade. The first is the
technology development of specialty fibers, especially double-clad rare-earth-doped silica-based optical fibers and the
second is the development of high power pump diodes. I will take this opportunity to review our work in the
development of specialty fibers and their applications in fiber lasers. I will try to cover three areas of development,
mode-area scaling using leakage channel fibers, highly ytterbium-doped low-photo-darkening silica-based fibers which
can be closely index-matched to silica, and silica suspended-core fibers for highly efficient super-continuum generations.
In recent years, a key approach for SBS suppression in fiber amplifiers for high power single frequency fiber lasers is
based on designs of transverse acoustic properties in fibers. Although this approach provides some SBS suppressions,
our new analysis considering the often omitted leaky acoustic modes demonstrates that it is fundamentally limited to SBS
threshold increase of less than ten-fold in the most ideal case. We have established an acoustic mode solver based on
simultaneous longitudinal and shear acoustic wave equations and rigorous boundary conditions. This enables us not only
the capability of studying designs of arbitrary acoustic velocity profiles, but also finding leaky acoustic modes with their
waveguide losses. In this work, we will demonstrate the applicability of acoustic modes and impact of leaky acoustic
modes in SBS in fibers, especially those with transverse acoustic velocity profiles designed to suppress SBS. With
consideration of the leaky acoustic modes, we have found that design of transverse acoustic velocity profiles only has a
very limited SBS suppression in fibers with even the most optimized acoustic anti-guides. Longitudinal acoustic property
profiling, on the other hand, can potentially provide well over two-orders of magnitude SBS suppression.
Traditional CARS microscopy using picosecond (ps) lasers has been applied to a wide variety of applications;
however, the lasers required are expensive and require an environmentally stable lab. In our work, we demonstrate
CARS microscopy using a single femtosecond (fs) laser combined with a photonic crystal fiber (PCF) and optimal
chirping to achieve similar performance to the ps case with important added advantages: fs-CARS utilizes
versatile source that allows CARS to be combined with other multiphoton techniques (e.g. SHG, TPF) for
multimodal imaging without changing laser sources. This provides an attractive entry point for many researchers
to the field. Furthermore, optimal chirping in fs-CARS also opens the door to the combination and extension
of other techniques used in ps CARS microscopy such as multiplex and FM imaging. The key advantage with
chirped fs pulses is that time delay corresponds to spectral scanning and allows for rapid modulation of the
resonant CARS signal. The combination of a fs oscillator with a PCF leads to a natural extension of the
technology towards an all-fiber source for multimodal multiphoton microscopy. An all-fiber system should be
more robust against environmental fluctuations while being more compact than free-space systems. We have
constructed and demonstrated a proof of concept all-fiber based source that can be used for simultaneous CARS,
TPF and SHG imaging. This system is capable of imaging tissue samples and live cell cultures with 4 μs/pixel
dwell time at low average powers.
Leakage channel fibers have demonstrated their ability to significantly extend the effective mode area of a fundamental
mode while maintaining robust single mode operation. These fibers are designed to have strong built-in mode filtering
which effectively suppresses the propagation of all higher order modes while keeping fundamental mode loss to a
minimum, and, therefore, effectively extending the regime of single mode operation. Recently all-glass leakage channel
fibers have been demonstrated as a significant improvement over designs with air holes. These all glass leakage channel
fibers not only can be manufactured with much improved consistence and uniformity. They can also be handled and used
as conventional fibers. More importantly, mode distortions from collapse of air holes in photonic crystal fibers during
splicing and other end face treatments are largely eliminated. We will review some of the recent progress in this area.
Critical power, nonlinear guided stationary mode and transient dynamics in nonlinear waveguides are studied
analytically. Under Gaussian mode approximation, critical power for nonlinear self-focus is derived analytically and is
found to be independent of waveguide parameters. Nonlinear guided stationary mode is found to be the stable solution of
nonlinear waveguide below critical power for nonlinear self-focus and has a reduced mode size dependent on optical
power and V value of the waveguide. Equation governing transient dynamics of the nonlinear waveguide modes is also
found. Transient dynamics scale with Rayleigh Raleigh range similar to that in bulk media. Mode is found to evolve
adiabatically towards the local stationary mode in a fiber below self-focus limit. Mode will collapse to a singular point at
a self-focus distance at and above critical power. Larger V value increases transient and self-focus distance. The self-focus
distance becomes proportional to square root of optical power at higher power level and independent of V value. B
integral are calculated for various amplifiers considering the impact of gradual collapse of beam size along the amplifier.
Leakage channel fibers, where several air holes form a core, can be precisely engineered to create large leakage loss for
higher order modes, while maintain negligible transmission loss for the fundamental mode. This unique property can be
used for designing optical fibers with large effective mode area, which supports robust fundamental mode propagation.
The large air holes in the design also enable the optical fibers to be bend-resistant. The principles of design and operation
regime are outlined in this paper. Performance of an ytterbium-doped double clad leakage channel fiber and an
ytterbium-doped polarization-maintaining (PM) double clad leakage channel fiber with ~50&mgr;m core diameter is also
Fiber laser technology has the potential to make a significant impact in many defense applications, from LIDAR and remote sensing to high energy laser weapons systems, in addition to numerous industrial, medical and scientific applications. This revolutionary new laser technology offers many intrinsic advantages over traditional DPSSLs and has received considerable support from key funding agencies over the last 10 years, including DARPA and the Joint Technology office (JTO) amongst others. With the aid of this funding several groups have now demonstrated small, compact and lightweight single mode fiber devices operating at the 1-10kW power level. There are research and development programs in the Air Force, Navy and Army currently. Funding to demonstrate beam combining of fiber lasers to the 25kW level is in place with programs such as RELI and Excalibur. More recently, the BAA call from the Office of Naval Research specifically calls for the development of high energy laser weapon technology suitable for ship board defense to be demonstrated on platform by 2016 and fiber technology is one possible solution being considered. Active deployment in the next decade is anticipated by some branches of the military.
Widespread publications in the research community have demonstrated an impressive array of power scaling fiber laser results, both CW and pulsed at wavelengths from 1um to the eyesafe 1.5um and 2um wavelengths. Advantages associated with fiber technology are not only high wallplug efficiency leading to reduced electrical power requirements and easier system cooling, but also robustness, good beam quality, compactness and highly flexible system performance. These, coupled with (remote) fiber delivery options make the technology unique in many applications.
This tutorial will cover the major aspects of designing and building a fiber laser, from the fiber itself through the various state of the art fiber components, and discussing the system parameter space that best makes use of the intrinsic advantages of the technology. Applications from industrial material processing though to defense and homeland security will be reviewed.