Contaminants can severely limit the efficiency, laser damage threshold, and strength of photonic crystal fiber-based lasers. Such contamination can occur due to environmental exposure during the pulling or stacking of rods and tubes or improper handling and storage of these glass components. A preform made by the “stack and draw” process is susceptible to incorporating surface contaminants into the bulk laser glass.
We have adapted cleaning and handling protocols originally developed for processing large fused silica optics for the National Ignition Facility. The etch cleaning process reported here mimics the “AMP” or “Advanced Mitigation Process” developed for NIF optics that see high fluence 351nm light. In addition, all cleaning, fixturing and assembly processes used to prep a stack for pulling into a fiber are done in a Class 100 cleanroom. Glass rods (1-3mm in diameter and 10” long) are assembled into a Teflon fixture that only contacts the rods at each end. The loaded fixture receives 120kHz ultrasonic cleaning in 10% sodium hydroxide at 45C and 3% Brulin 1696 detergent at 55C. Parts are thoroughly rinsed using ultrasonicated ultrapure water and spray rinses. A 200nm etch in buffered hydrofluoric acid (6:1 BOE diluted 2:1 in DI water) is followed by additional ultasonicated (120kHz-270kHz) ultrapure water and spray rinse. Finally, the components are allowed to fully dry inside the Teflon frame. The rods are cleaned, stacked, and assembled into a fused silica tube.
The preform stack is then returned to a non-cleanroom area to be pulled into fiber using standard telecom fiber-based draw tower equipment and without clean air filters around the draw area. Four fibers were made to test independently the damage threshold and the background loss, two Yb core active fibers and two silica core (F clad) fibers. One of each was cleaned with the AMP process, and one of each with a methanol wipe cleaning process. The active fiber was coated with a dual acrylate coating, first with a low-index inner coating to provide a pump cladding, and then with a relatively hard coating to protect the relatively soft primary coating. The active fibers were pumped at 980nm in a double Fresnel cavity configuration and the power increased until the fiber was damaged up to 1kW. The passive fiber background loss was measured using a standard cut-back technique. Replacing the former methanol wipe clean process with this aqueous cleaning process improved the 1060nm damage threshold of a fiber laser by >30x to above the kW level in the laboratory and reduced the background attenuation by >18x. Early indications are that the acid etching also makes the tensile strength of the fiber consistently high.
This work was performed under the auspices of the U.S. Department of Energy by Lawrence
Livermore National Laboratory under Contract DE-AC52-07NA27344.
The performance of high-contrast AO instruments (GPI, SPHERE, ScEXAO, MagAO) and other systems that operate at visible wavelengths can be severely hampered by control system latencies and temporal wavefront errors. In high-contrast systems, temporal errors and delays are manifest as high spatial frequency wavefront residuals that scatter light into the controllable region of the PSF and diminish contrast, an effect that is particularly severe when atmospheric coherence times are short. Solutions that have been proposed include lower latency electronics, deformable mirrors with lower mechanical response times, and specialized control algorithms such as predictive control. These advancements will be necessary for achieving the latency goals of high actuator count systems on future Extremely Large Telescopes (ELTs), including NFIRAOS+ and PFI on the Thirty Meter Telescope, upgrading the performance of existing highcontrast systems, and pushing adaptive optics to visible wavelengths. LLAMAS (Low-Latency Adaptive Optical Mirror System) is a fully funded adaptive optics system at the Lawrence Livermore National Laboratory site that will test these techniques in an integrated, real time, closed-loop AO system. With a total system latency goal of ~100 microseconds (including mechanical response time, not including frame integration), LLAMAS will achieve an order of magnitude improvement in AO system latencies over the current generation of high-contrast AO systems. The woofer/tweeter architecture will incorporate a 492-actuator Boston Micromachines MEMS device mapping 24 actuators across a circular pupil. The tweeter mirror will be paired with a specialized low-latency driver, delivering less than 40 microseconds electronic and mechanical latency (10 – 90%). The real-time control computer will utilize the computationally efficient Fourier Transform Reconstructor with a predictive Kalman filter with a goal of completing all computations and reconstructing the wavefront in less than 20 microseconds. LLAMAS will be fully integrated with a 21×21 lenslet Shack-Hartmann sensor by January 2019. These proceedings describe the LLAMAS design, characterize the performance of its low-latency componentry, and discuss the relevance of the design for future high-contrast, visiblelight, and high actuator count AO systems on ELTs.
We present 10W single-mode fiber laser based on Nd+3 fiber operating at 1428nm. All-solid fused silica microstructured waveguide fiber design is employed to suppress amplification at 1μm. The Nd+3 fiber is pumped by commercial multi-mode 880nm diode.
This talk will provide an overview of high power laser research at Lawrence Livermore National Laboratory (LLNL). It will discuss the status of the National Ignition Facility (NIF) laser. In addition, the talk will describe other laser development activities such as the development of high average power lasers and novel fiber lasers.
The National Ignition Facility (NIF) has been in service since 2007 and operating with > 1 MJ energies since 2009. During this time the facility has transitioned to become an international user facility and increased the shot rate from ~150 target shots per year to greater than 400 shots per year. Today, the NIF plays an essential role in the US Stockpile Stewardship Program, providing data under the extreme conditions needed to validate computer models and train the next generation of stockpile stewards. Recent upgrades include the Advanced Radiographic Capability (ARC), a high energy short pulse laser used to do high resolution radiography.
In addition to the NIF, this talk will include an overview of progress on the high average power laser development, recent results from fiber laser development activities and improvements to laser design and computational capabilities.
Power scaling using a higher order mode in a ribbon fiber has previously been proposed. However, methods of selecting the higher order mode and converting to a single lobe high brightness beam are needed. We propose using a multiplexed transmitting Bragg grating (MTBG) to convert a higher order mode into a single lobe beam. Using a ribbon fiber with core dimensions of 107.8 μm by 8.3 μm, we use the MTBG to select a higher order mode oscillating within the resonator with 51.4% efficiency, while simultaneously converting the higher order mode to a beam with diffraction limited divergence of 10.2 mrad containing 60% of the total power.
We have investigated stimulated Raman scattering in the 4H polytype of SiC, due to its excellent thermal conductivity which is of great importance for power scaling of Raman lasers. Spectroscopy verifies the sample’s polytype and precludes any significant admixture of other polytypes. Tests indicate the moderate optical quality of this commercially available sample. Using pump-probe measurements around 1030 nm, we find the Raman gain coefficient of the major peak at 777 cm-1 to be 0.46 cm/GW. Although this value is only modest, calculations and experience with other Raman materials indicate that Raman lasing of 4H SiC should be possible with reasonable intensities of 1064-nm pulsed pumping.
Increasing the dimensions of a waveguide provides the simplest means of reducing detrimental nonlinear effects, but such systems are inherently multi-mode, reducing the brightness of the system. Furthermore, using rectangular dimensions allows for improved heat extraction, as well as uniform temperature profile within the core. We propose a method of using the angular acceptance of a transmitting Bragg grating (TBG) to filter the fundamental mode of a fiber laser resonator, and as a means to increase the brightness of multi-mode fiber laser. Numerical modeling is used to calculate the diffraction losses needed to suppress the higher order modes in a laser system with saturable gain. The model is tested by constructing an external cavity resonator using an ytterbium doped ribbon fiber with core dimensions of 107.8μm by 8.3μm as the active medium. We show that the TBG increases the beam quality of the system from M2 = 11.3 to M2 = 1.45, while reducing the slope efficiency from 76% to 53%, overall increasing the brightness by 5.1 times.
The National Ignition Facility (NIF) at Lawrence Livermore National Laboratory is the first of a kind megajoule-class laser with 192 beams capable of delivering over 1.8 MJ and 500TW of 351nm light , . It has been commissioned and operated since 2009 to support a wide range of missions including the study of inertial confinement fusion, high energy density physics, material science, and laboratory astrophysics. In order to advance our understanding, and enable short-pulse multi-frame radiographic experiments of dense cores of cold material, the generation of very hard x-rays above 50 keV is necessary. X-rays with such characteristics can be efficiently generated with high intensity laser pulses above 1017 W/cm² . The Advanced Radiographic Capability (ARC)  which is currently being commissioned on the NIF will provide eight, 1 ps to 50 ps, adjustable pulses with up to 1.7 kJ each to create x-ray point sources enabling dynamic, multi-frame x-ray backlighting. This paper will provide an overview of the ARC system and report on the laser performance tests conducted with a stretched-pulse up to the main laser output and their comparison with the results of our laser propagation codes.
Power scaling of high power laser resonators is limited due to several nonlinear effects. Scaling to larger mode areas can offset these effects at the cost of decreased beam quality, limiting the brightness that can be achieved from the multi-mode system. In order to improve the brightness from such multi-mode systems, we present a method of transverse mode selection utilizing volume Bragg gratings (VBGs) as an angular filter, allowing for high beam quality from large mode area laser resonators. An overview of transverse mode selection using VBGs is given, with theoretical models showing the effect of the angular selectivity of transmitting VBGs on the resonator modes. Applications of this ideology to the design of laser resonators, with cavity designs and experimental results presented for three types of multimode solid state lasers: a Nd:YVO4 laser with 1 cm cavity length and 0.8 mm diameter beam with an M2 of 1.1, a multimode diode with diffraction limited far field divergence in the slow axis, and a ribbon fiber laser with 13 cores showing M2 improved from 11.3 to 1.5.
Diffraction-limited high power lasers in the region of 10s of kW to greater than 100 kW are needed for defense, manufacturing and future science applications. A balance of thermal lensing and Stimulated Brillouin Scattering (SBS) for narrowband amplifiers and Stimulated Raman Scattering (SRS) for broadband amplifiers is likely to limit the average power of circular core fiber amplifiers to 2 kW (narrowband) or 36 kW (broadband). A ribbon fiber, which has a rectangular core, operating in a high order mode can overcome these obstacles by increasing mode area without becoming thermal lens limited and without the on-axis intensity peak associated with circular high order modes. High order ribbon fiber modes can also be converted to a fundamental Gaussian mode with high efficiency for applications in which this is necessary. We present an Yb-doped, air clad, optical fiber having an elongated, ribbon-like core having an effective mode area of area of 600 μm² and an aspect ratio of 13:1. As an amplifier, the fiber produced 50% slope efficiency and a seed-limited power of 10.5 W, a gain of 24 dB. As an oscillator, the fiber produced multimode power above 40 W with 71% slope efficiency and single mode power above 5 W with 44% slope efficiency. The multimode M2 beam quality factor of the fiber was 1.6 in the narrow dimension and 15 in the wide dimension.
Diffraction limited fiber amplifiers in a circular geometry are likely to be limited by nonlinearities to 2 kW for narrowband and 10-36 kW for broadband lasers. We have proposed a ribbon fiber geometry to allow scaling fiber lasers above these limits in which a high order ribbon mode is amplified and converted back to the fundamental mode in free space. Novel methods of illuminating a high order ribbon fiber mode are discussed and compared with modeling and experimental results showing high purity illumination, > 90%. A 10 kW single frequency ribbon fiber amplifier design is presented and BPM simulation results verify the approach.
A rectangular-core fiber that guides and amplifies a
higher-order-mode can potentially scale to much higher
average powers than what is possible in traditional circular core large-mode-area fibers. Such an amplifier
would require mode-conversion at the input and output to enable interfacing with TEM00 mode seed sources
and generate diffraction-limited radiation for various applications. We discuss the simulation and experimental
results of a mode conversion technique that uses two
diffractive-optic-elements in conjugate Fourier planes to
convert a diffraction limited TEM00 mode to the higher-order-mode of a ribbon core fiber. Our experiments
show that the mode-conversion-efficiency exceeds 84% and can theoretically approach 100%.
A developed formalism1 for analyzing the power scaling of diffraction limited fiber lasers and amplifiers is applied to a
wider range of materials. Limits considered include thermal rupture, thermal lensing, melting of the core, stimulated
Raman scattering, stimulated Brillouin scattering, optical damage, bend induced limits on core diameter and limits to
coupling of pump diode light into the fiber. For conventional fiber lasers based upon silica, the single aperture,
diffraction limited power limit was found to be 36.6kW. This is a hard upper limit that results from an interaction of the
stimulated Raman scattering with thermal lensing. This result is dependent only upon physical constants of the material
and is independent of the core diameter or fiber length. Other materials will have different results both in terms of
ultimate power out and which of the many limits is the determining factor in the results. Materials considered include
silica doped with Tm and Er, YAG and YAG based ceramics and Yb doped phosphate glass. Pros and cons of the
various materials and their current state of development will be assessed. In particular the impact of excess background
loss on laser efficiency is discussed.
We have previously demonstrated ultrashort pulse amplification in fiber systems beyond the B-integral limit. Here we
report on recent experiments to increase the average power of such systems, and their application to high-speed material
processing. A compact fiber chirped pulse amplification system, producing sub-picosecond 50 &mgr;J pulses at a repetition
rate of 1 MHz, is obtained by implementing a fiber stretcher and a 1780 l/mm dielectric diffraction grating compressor.
Despite a substantial residual dispersion mismatch between stretcher and compressor, the cubicon fiber amplifier allows
for the generation of sub-picosecond pulses with sufficient quality for high-speed micromachining applications.
Moreover, the dielectric grating compressor allows power independent near-diffraction limited beam quality as required
for precision micro-machining. We utilize this laser to mill aluminum, alumina, and glass targets with material removal
rates >0.2 mm3/s in all three materials.
We have demonstrated 3.5W of 589nm light from a fiber laser using periodically poled stoichio-metric Lithium Tantalate (PPSLT) as the frequency conversion crystal. The system employs 938nm and 1583nm fiber lasers, which were sum-frequency mixed in PPSLT to generate 589nm light. The 938nm fiber laser consists of a single frequency diode laser master oscillator (200mW), which was amplified in two
stages to >15W using cladding pumped Nd3+ fiber amplifiers. The fiber amplifiers operate at 938nm and minimize amplified spontaneous emission at 1088nm by employing a specialty fiber design, which maximizes the core size relative to the cladding diameter. This design allows the 3-level laser system to operate at high inversion, thus making it competitive with the 1088nm 4-level laser transition. At 15W, the 938nm laser has an M2 of 1.1 and good polarization (correctable with a quarter and half wave plate to >15:1). The 1583nm fiber laser consists of a Koheras 1583nm fiber DFB laser that is pre-amplified to 100mW, phase modulated and then amplified to 14W in a commercial IPG fiber amplifier. As a part of our research efforts we are also investigating pulsed laser formats and power scaling of the 589nm system. We will discuss the fiber laser design and operation as well as our results in power scaling at 589nm.
We are developing an all fiber laser system optimized for providing input pulses for short pulse (1-10ps), high energy (~1kJ) glass laser systems. Fiber lasers are ideal solutions for these systems as they are highly reliable and enable long term stable operation. The design requirements for this application are very different than those commonly seen in fiber lasers. High-energy lasers often have low repetition rates (as low as one pulse every few hours), and thus high average power and efficiency are of little practical value. What is of high value is pulse energy, high signal to noise ratio (expressed as pre-pulse contrast), good beam quality, consistent output parameters and timing. Our system focuses on optimizing these
parameters. Our prototype system consists of a mode-locked fiber laser, a compressed pulse fiber amplifier, a "pulse cleaner", a chirped fiber Bragg grating, pulse selectors, a transport fiber system and a large mode area fiber amplifier. We will review the system and present theoretical and experimental studies of critical aspects, in particular the requirement for high pre-pulse contrast.
We have tested a series of Ytterbium doped large core fibers operating near 10Kpps and producing pulses of approximately 1ns. We have achieved 0.85mJ/pulse resulting in peak powers in excess of 2MW with 0.4ns pulses and near diffraction limited beams. In another fiber, we have achieved over 1.5mJ/pulse with pulses of 900ps corresponding to 1.65MW of peak power and M2 of 2.5. In the latter case, wall-plug efficiencies, excluding cooling of the pump diode lasers, in excess of 15% were also achieved. This fiber amplifier has operated for 2 months without any degradation or observed optical damage.
We have demonstrated 466 mW of 469 nm light from a frequency doubled continuous wave fiber laser. The system consisted of a 938 nm single frequency laser diode master oscillator, which was amplified in two stages to 5 Watts using cladding pumped Nd 3+ fiber amplifiers and then frequency doubled in a single pass through periodically poled KTP. The 3 cm long PPKTP crystal was made by Raicol Crystals Ltd. with a period of 5.9 μm and had a phase match temperature of 47 degrees Centigrade. The beam was focused to a 1/e2 diameter in the crystal of 29 μm. Overall conversion efficiency was 11% and the results agreed well with standard models. Our 938 nm fiber amplifier design minimizes amplified spontaneous emission at 1088 nm by employing an optimized core to cladding size ratio. This design allows the 3-level transition to operate at high inversion, thus making it competitive with the 1088 nm 4-level transition. We have also carefully chosen the fiber coil diameter to help suppress propagation of wavelengths longer than 938 nm. At 2 Watts, the 938 nm laser had an M2 of 1.1 and good polarization (correctable with a quarter and half wave plate to >10:1).
We are developing an all fiber laser system optimized for providing input pulses for short pulse (1-10ps), high energy (~1kJ) glass laser systems. Fiber lasers are ideal solutions for these systems as they are highly reliable and once constructed they can be operated with ease. Furthermore, they offer an additional benefit of significantly reduced footprint. In most labs containing equivalent bulk laser systems, the system occupies two 4’x8’ tables and would consist of 10's if not a 100 of optics which would need to be individually aligned and maintained. The design requirements for this application are very different those commonly seen in fiber lasers. High energy lasers often have low repetition rates (as low as one pulse every few hours) and thus high average power and efficiency are of little practical value. What is of high value is pulse energy, high signal to noise ratio (expressed as pre-pulse contrast), good beam quality, consistent output parameters and timing. Our system focuses on maximizing these parameters sometimes at the expense of efficient operation or average power. Our prototype system consists of a mode-locked fiber laser, a compressed pulse fiber amplifier, a “pulse cleaner”, a chirped fiber Bragg grating, pulse selectors, a transport fiber system and a large flattened mode fiber amplifier. In our talk we will review the system in detail and present theoretical and experimental studies of critical components. We will also present experimental results from the integrated system.
Continuous wave (CW) fiber laser systems with output powers in excess of 500 W with good beam quality have now been demonstrated, as have high energy, short pulse, fiber laser systems with output energies in excess of 1 mJ. Fiber laser systems are attractive for many applications because they offer the promise of high efficiency, compact, robust systems. We have investigated fiber lasers for a number of applications requiring high average power and/or pulse energy with good beam quality at a variety of wavelengths. This has led to the development of a number of custom and unique fiber lasers. These include a short pulse, large bandwidth Yb fiber laser for use as a front end for petawatt class laser systems and a narrow bandwidth 0.938 μm output Nd fiber laser in the > 10 W power range.
We report initial operation of the Mercury laser with seven 4 x 6 cm S-FAP amplifier slabs pumped by four 80 kW diode arrays. The system produced up to 33.5 J single shot, 23.5 J at 5 Hz, and 10 J at 10 Hz for 20 minute runs at 1047 nm. During the initial campaign, more than 2.8 x 104 shots were accumulated on the system. The beam quality of the system was measured to be 2.8 x 6.3 times diffraction limited at 110 W of output, with 96% of the energy in a 5X diffraction limited spot. Static wavefront glass plates were used to correct for the low order distortions in the slabs due to fabrication and thermal loading. Scaling of crystal grown has begun with the first full size slab produced from large diameter growth. Using an energetics optimization code we find the beam aperture is scalable up to 20 x 30 cm and 4.2 kJ.
We have developed and demonstrated a large flattened mode (LFM) optical fiber, which raises the threshold for non-linear interactions in the fiber core by a factor of 2.5 over conventional large mode area fiber amplifiers. The LFM fiber works by incorporating a raised index ring around the outer edge of the fiber core, which serves to flatten the fundamental fiber mode from a Bessel function to a top hat function. This increases the effective area of the core intersected by the mode by a factor of 2.5 without increasing the physical size of the core. This is because the core is uniformly illuminated by the LFM mode rather than having most of the light confined to the center of the core. We present experimental and theoretical results relating to this fiber and its design.
We have developed a Nd:doped cladding pumped fiber amplifier, which operates at 938nm with greater than 2W of output power. The core co-dopants were specifically chosen to enhance emission at 938nm. The fiber was liquid nitrogen cooled in order to achieve four-level laser operation on a laser transition that is normally three level at room temperature, thus permitting efficient cladding pumping of the amplifier. Wavelength selective attenuation was induced by bending the fiber around a mandrel, which permitted near complete suppression of amplified spontaneous emission at 1088nm. We are presently seeking to scale the output of this laser to 10W. We will discuss the fiber and laser design issues involved in scaling the laser to the 10W power level and present our most recent results.
We report on the frist experimental demonstration of a scalable fiber laser approach based on phase-locking multiple gain cores in an antiguided structure. A novel fabrication technology is used with soft glass components to construct the multie core fiber used in our experiments. The waveguide region is rectangular in shape and comprised of a periodic sequence of gain and no-gain segments having nearly uniform refractive index. The rectangular waveguide is itself embedded in a lower refractive index cladding region. Experimental resutls confirm taht our five-core Nd doped glass prototyep structure runs predominately in two spatial antiguided modes as predicted by our modeling.
Bragg Gratings are waveguides, typically single-mode optical fibers, into which a periodic refractive index modulation has been imprinted by a patterned UV exposure. Fiber Bragg Gratings separate telecom frequency bands or compensate for optical dispersion in long-haul fiber networks, and also serve as strain sensors for civil engineering or geophysical studies and oil, gas or mining exploitation. A Bragg Grating writer is an interferometer for generating the UV exposure pattern. It is one of the unusual cases where an interferometer is a production tool, rather than a metrology instrument. In this paper, we review the most common Bragg Grating writing geometry and propose an opto-mechanical structure having minimal adjustment and very high mechanical stability.
A detailed overview of the design of an optical current sensor implemented with all fiber optic components is presented. Laboratory and field data representing stability with time and temperature is included. Fundamental limits and their relationship to product specifications are also discussed.