We present the ELAN testbed, which aims at assessing laser-induced damage of thick and large silica optics for the Laser Megajoule (LMJ) facility. A custom and patented injection seeding system was implemented on our Nd:YAG laser: this allows us to use phase modulated pulses to annihilate backward stimulated Brillouin scattering during tests while guaranteeing a smooth temporal profile. Then, beam propagation simulations of the incident spatial profile with Miro code (CEA) allows us to correct for the surface narrowing of the beam on the exit face of the optic due to Kerr self-focusing. This testbed allows us to test LMJ 34 mm vacuum window at 351 nm in nanosecond range with fluence up to 30-35 J/cm² with a control of non-linear propagation effect.
Over the past two decades, laser technology has made significant strides in scaling peak and average power levels. These advancements span a wide range of laser irradiation techniques, from ultrashort pulses to continuous waves, driving the development of various optical elements and coatings. However, the variability in optics sizes, ranging from optical fibers to meter-sized optics, and the presence of various failure modes pose challenges for laser damage testing. Consequently, there is a pressing need to align relevant laser damage testing standards to ensure the functional performance of optics. In this overview, we discuss recent standards-revision efforts aimed at revising the ISO 21254 family standards. Our primary focus is on improving accuracy and reliability by improving damage criteria, testing procedures, and results analysis methods. These efforts aim to tackle emerging challenges in laser damage testing while ensuring that standards remain compatible with modern technological developments.
ISO standards are periodically reviewed to ensure their relevance to the best industry practices. Significant advancements have been observed in laser source development and related technologies over the last two decades. These advancements encompass new irradiation regimes, ranging from ultrashort pulses to kW-class continuous wave irradiation, with substantially increased peak- and average laser power levels. This new reality also necessitates the adaptation of pertinent laser damage testing standards. As high-power laser applications introduce optical elements with unique failure mechanisms and size constraints, there is a growing need for the introduction of alternative testing methods. In this paper, we provide a brief overview of recent standardization efforts undertaken by ISO TC 172 SC 9 WG 1 for the revision of the ISO 21254 series standards - ”Lasers and laser-related equipment — Test methods for laser-induced damage threshold”. Specifically, we discuss the need for the extension of ’classical’ damage criteria, the introduction of alternative test procedures, and possible improvements in interrogation methods and analysis. The overarching goal of this paper is to promote transparency in the standardization process and inspire discussion, ultimately leading to the enhancement of accuracy and reliability in laser damage testing.
Accurate laser damage testing for Laser MégaJoule (LMJ) fused silica optics is crucial for predicting their lifetimes. However, beam propagation in optics is usually neglected and yet, damage sites are mostly initiated on the exit face of optics. The MELBA testbed in CEA CESTA (France) delivers nanosecond UV laser pulses representative of LMJ optics conditions on a centimeter scale. Our particular imaging system measures the beam spatial profile before and after propagation in samples, enabling quantification of self-focusing induced by the Kerr effect. This metrology is necessary for laser damage parametric studies, particularly when these laser parameters influence the Kerr effect. We present here a study of the impact of a linear-to-circular polarization conversion on laser damage, which highlights the importance of accurately assessing non-linear beam propagation for laser damage tests.
We investigate the feasibility of a full-silica transmission grating acting as a quarter-wave plate at the wavelength of 351 nm as an alternative to an anisotropic crystal. We report on the design, manufacturing process and optical characterization carried out. We evidence the possibility to obtain a full-silica component exhibiting at 351 nm a high damage threshold and a phase retardance of /4 associated with a diffraction efficiency above 98%.
The poster introduces the MELBA setup located at CEA CESTA (France). The MELBA laser delivers a nanosecond UV centimeter-sized laser beam and is dedicated to the study of laser-induced damage and damage growth within the Laser MégaJoule framework. Laser pulses are spatially, temporally and spectrally both shaped and characterized. A dedicated imaging system can measure the non-linear propagation in samples and its consequence on surface damage and filamentation. Recently, it was made possible to adjust the beam polarization from linear to circular.
High-energy laser pulses in the nanosecond regime used to be spectrally broadened to mitigate the stimulated Brillouin scattering known to deteriorate the optical elements. Due to propagating effects, this spectrum broadening lead to FM-to-AM conversion, where the UV laser beam experiences an amplitude modulation at frequencies which are harmonics of the phase modulation frequency. We study the impact of the FM-to-AM conversion on the Brillouin backscattering by applying an amplitude modulation on the UV pump laser beam operating at 351 nm and with a 3 ns pulse duration.
Experimental measurements show that adding an amplitude modulation frequency on a phase-modulated laser beam could enhance the stimulated Brillouin scattering and lead to laser damage. Thanks to a theoretical and numerical analysis, we show that this singular behavior originates from a resonance between the frequency of the amplitude modulation and the low orders harmonic frequencies of the phase modulated laser beam.
Possible linear-to-circular polarization conversion had been studied for the Laser MégaJoule. We measured the consequences of such polarization conversion on laser-induced damage using the MELBA testbed. The MELBA laser is located in CEA CESTA (France) and delivers a nanosecond UV centimeter-sized laser beam. Experimental comparison of polarizations states showed a significant decrease of damage densities in circular polarization. Thanks to the particular imaging setup, we were able to explain this by both a reduction of the Kerr effect (supported by theory) and a reduction of the intrinsic absorption of silica optics defects.
Multilayer dielectric (MLD) gratings provide high diffraction efficiency and a high damage threshold. They represent the main solution to compressing a high-power laser beam. However, the laser resistance of MLD gratings limits the power of such facilities. The community devoted a lot of resources to increasing the damage threshold of those components. Today, it is well known that the etching profile plays a key role in the electric field distribution and consequently the laser resistance. In this paper, we focused our optimization on the multilayer dielectric stack to increase the laser-induced damage threshold (LIDT). We numerically and experimentally demonstrated the impact of the MLD stack on the electric field distribution and the LIDT. We manufactured two sets of three samples with identical etching profiles. The calculated electric field intensities were in good agreement with the measured LIDTs. These results demonstrated how to further optimize grating designs through the dielectric stack.
Laser MegaJoule (LMJ) is a high energy laser facility designed for fusion experiments. To track final optics damage, laser damage monitoring is carried out using images acquired by a camera. To prepare for the LMJ full energy/power operation, the damage models based on the phenomenological laws established in the laboratories are validated by experimental campaigns dedicated to performance. The very high quality of LMJ optics surfaces makes damage highly unlikely. In order to take the greater benefit of these performance campaigns, carried out on a reduced number of laser shots and components, a matrix of nearly 1000 damage sites is initiated offline on one optics. Precisely measured on a metrology bench before and after the campaign, this component was on LMJ facility during a performance campaign at the end of 2021. Very useful for the calibration of the LMJ monitoring camera, it also provided data to set LMJ laser damage models at higher energy level.
The poster will report the Laser Induced Damage Threshold of an amplifier medium immersed in an active cooling system, and with different protocols, as 1on1 and Son1 procedures to evaluate the influence of the cooling system on the laser damage resistance. Some theoretical investigations will also be reported in order to explain the different experimental observations.
The MELBA testbed located at CEA CESTA (France) is a nanosecond UV centimeter-sized beam aiming at studying laser damage. Previous work carried out on MELBA showed a dependency of laser-induced damage to power modulation of the laser pulse in linear propagation regime. We extended this study to the non-linear propagation regime. Thanks to dedicated imaging setup, we carried out an experimental study of Bespalov-Talanov gain and laser-induced damage with different temporal modulations. The tuning of phase modulation parameters made it possible to annihilate backward stimulated Brillouin scattering and consequently focus on the impact of the Kerr effect.
The paper aims to compare the results obtained with the same laser source with a large beam and with small beams. These
latest were shaped from phase objects implemented to obtain several small beams from a single larger beam. The
consistency of the results from both sets of measurements is shown. It validates the assumptions made and the specific
mathematical treatments implemented to establish the link between the two approaches. It also validates and strengthens
the approach developed from the rasterscan procedure used to measure damage densities from the scanning with beams
of small dimensions. This shows that small beam tests are reasonably representative of tests carried out with large beams.
Multilayer dielectric gratings (MLDG) are key optical components of Petawatt-class laser that are used to compress short pulses of high intensities. Laser-induced damage can occur on the top area of the components, typically arising in the pillars periodically etched. This phenomenon limits the power yielded by high power laser facilities such as PETAL (PETwatt Aquitaine Laser) laser facility. PETAL is expected to delivery pulses with a wavelength around 1053 nm, an energy around 3 kJ and a pulse duration between 0.5 and 10 ps. Coupled with LMJ (Laser MegaJoule), PETAL aims to study materials in extreme conditions to reproduce the environment in the heart of stars or planets, fusion by inertial confinement, particularly rapid ignition and shock ignition, and nuclear physics for medical proton therapy. In this study, we present a process to improve the laser-induced damage threshold of PETAL pulse-compression gratings in sub-picosecond regime by reducing the electric field intensity in the pillars. PETAL gratings have specific parameters of operation: Transverse Electric polarization, under vacuum, a period equal to 1780 lines per mm and diffraction efficiency higher than 95% for the -1st order. Theoretical designs are calculated with a code developed at the Fresnel Institute. The code solves Fresnel equations by using the differential method, Fast Fourier Factorization (FFF) and S matrix propagation algorithm. As a result, we obtain the distribution of the electric field and diffraction efficiency of any given diffraction order. First, starting with a given MLD mirror, we calculate an etching profile that maximizes the diffraction efficiency at the -1st order by taking into account the manufacturing constraints of future suppliers. Then, we optimize the mirror stack without changing the etching profile. We modify only the first top layers under the grooves. We obtained theoretical designs with the same etching profile and identical diffraction efficiency, associated with different electric field intensity values and expected different laser induced damage thresholds.
Final fused silica optics of high energy fusion class laser facilities are one of the components that limit the UV laser energy available for experiments. These final optics suffer from laser-induced damage. Some solutions are available to limit laser damage growth and to increase optics lifetime. However, to use them, it is necessary to be able to detect damage initiation as soon as possible, and to follow damage growth efficiently. An imaging system and a lighting source make the observation of laser damage sites possible after each laser shot without removing the optical components. Laser damage detection algorithms exist but they are not sufficiently efficient to provide reliable monitoring of damage growth over time because of small repositioning fluctuations of the optical system. An effective solution based on digital image correlation and brightness/contrast corrections is proposed to detect and follow laser damage sites as soon as they initiate in an automatic way. The effectiveness of the presented method is compared to the widely used method that is based on the analysis of local signal-to-noise ratio.
The standardization and comparison of laser-damage protocols and results are essential prerequisites for development and quality control of large optical components used in high-power laser facilities. To this end, the laser-induced–damage thresholds of two different coatings were measured in a round-robin experiment involving five well-equipped damage testing facilities. Investigations were conducted at the wavelength of 1 μm in the sub picosecond pulse duration range with different configurations in terms of polarization, angle of incidence, and environment (air versus vacuum). In this temporal regime, the damage threshold is known to be deterministic, i.e., the continuous probability distribution transitions from 0 to 1 over a very narrow fluence range. This in turn implies that the damage threshold can be measured very precisely. These characteristics enable direct comparison of damage-threshold measurements between different facilities, with the difference in the measured values indicating systematic errors or other parameters that were not previously appreciated. The results of this work illustrate the challenges associated with accurately determining the damage threshold in the short-pulse regime. Specifically, the results of this round-robin damage-testing effort exhibited significant differences between facilities. The factors to be taken into account when comparing the results obtained with different test facilities are discussed: temporal and spatial profiles, environment, damage detection, sample homogeneity, and nonlinear beam propagation.
The paper aims to compare the results obtained with the same laser source with a large beam and with small beams. These latest were shaped from phase objects implemented to obtain several small beams from a single larger beam. The consistency of the results from both sets of measurements is shown. It validates the assumptions made and the specific mathematical treatments implemented to establish the link between the two approaches. It also validates and strengthens the approach developed from the rasterscan procedure used to measure damage densities from the scanning with beams of small dimensions. This shows that small beam tests are reasonably representative of tests carried out with large beams.
The standardization and the comparison of laser-damage testing are essential prerequisites for development and quality control of large optical components used in high-power laser facilities. To this end, the laser-induced–damage thresholds of two different coatings were measured at four laboratories involved in a round-robin experiment. Tests were conducted at 1 m in the subpicosecond range with different configurations in terms of polarization, angle of incidence, and environment (air versus vacuum). In this temporal regime, the damage threshold is known to be deterministic, i.e., the continuous probability distribution transitions from 0 to 1 over a very narrow fluence range. This in turn implies that the damage threshold can be measured very precisely. These traits enable direct comparison of damage-threshold measurements between different facilities, while the difference in the measured values are not accompanied by large statistical uncertainties.
In this presentation, the results of this comparative experiment are compiled, illustrating the challenges associated with accurately determining the damage threshold in the short-pulse regime. Specifically, the results of this this round-robin damage-testing effort exhibited significant differences between facilities. The factors to be taken into account when comparing the results obtained with different test facilities are discussed: temporal and spatial profiles, environment, damage detection, samples homogeneity and nonlinear beam propagation.
We report on our recent work on the effect of defects on laser damage initiation and growth in the ultrashort regime. We particularly investigate the mechanisms of energy deposition in nano / micro scale defects in the coatings and their relation to damage initiation and subsequent damage growth using different experimental tools such as in-situ observation of damage developments, pump probe-microscopy, and studies on engineered model defects.
Laser damage resistance is a key factor for the operation and the improvement of high power laser systems. Up today laser damage performance of optical components is mainly a defect related material characteristic. Metrology procedures have been developed to realize repeatable and accurate measurements of surface damage density due to nanosecond pulses. These measurement techniques were used to guide the improvement of surface damage resistance. Fractures must be eliminated from surfaces, in order not to suffer a damage growth phenomenon, whose exponential character will reduce the optical lifetime. High intensity hot spots due to beam modulations, spatial and/or temporal modulations, can also cause surface damage. Specific set-ups and experiments were carried out that allowed us to analyze and explain these phenomena: damage initiation mechanisms, damage growth and beam propagation inside the optical components. The presentation aims to highlight relevant progress on these topics (1) initiation mechanisms due to defects show two phases: a first incubation phase followed by the expansion one of the damage site ; it appears also that damage diameters are well correlated with the expansion fluence, (2) a refined bulk observation coupled to a fractal analysis allow the quantification of bulk damage and therefore to explain the laser damage growth and its dependence on beam parameters like pulse length and fluence (3) non-linear amplification of phase and amplitude modulations lead to the amplification of the spatial pre-existing small scale modulations that increase the laser energy density locally and finally the number of damage sites. The whole of results, damage initiation, damage growth and beam propagation, is discussed to the light of the laser damage observed on LMJ optics.
Laser damage resistance is a key factor for the improvement of high power laser system. The PETAL laser, developed by the CEA-CESTA (France), uses meter scale reflective optics to compress, transport and focalize sub-picosecond laser pulses at 1053nm with high-energy [1]. In the case of defect-free material, laser-induced damage in the sub-picosecond regime is known to be deterministic since the threshold depends only on the electronic structure of the irradiated materials, the pulse duration and the enhancement of the electric fields in thin film coatings. Based on this consideration, a mono-shot technique has been investigated to assess the intrinsic damage resistance of optical component with only one laser shot. On the other hand, while considering real optical components, manufacturing processes included nanoscale defects in the functional coating. These defects can be ejected when irradiated and strongly reduce the laser damage resistance of optics: rasterscan procedure has then been developed to determine defect-induced damage densities. These densities are found to be high even for fluences well below the intrinsic Laser-Induced Damage Threshold and they increase with the fluence. These experiments bring new information on the operating characteristics of optics in short pulse regime. Once damage is triggered, its evolution under subsequent irradiations has also been studied. Growth experiments have been compared to numerical simulations. The investigations on growth behavior allow a better estimation of the functional lifetime of an optic in its operating conditions. The whole of results, damage initiation and damage growth, is discussed to the light of the laser damage observed on PETAL optics.
A millimetric aperture Nd:glass laser system has been designed and constructed at the CEA-CESTA. Its aim is to best mimic the laser conditions that can be found in inertial confinement fusion facilities. It is therefore used to study the main phenomena that prevents these lasers to work at their maximum power: the laser induced damage of the optical components. The combination of temporal and spatial modulators provides, every minute, a 6 J, 7 mm, 351 nm homogeneous beam at the fused silica sample location. This proceeding illustrates the capacity of the facility over two experiments: the study of damage initiation and the growth of laser damage sites on fused silica, up to millimetric scales
The growth of damage sites from micrometric to millimetric scales under high energy laser system conditions
have herein been investigated. In this realm, a saturation of the surface growth followed by the rapid expansion of
radial cracks has been observed. This observation contrasts with the previously reported exponential behavior1
(for pulse durations above 2 ns) and linear behavior (for pulse durations below 2 ns). The observation of the
longitudinal damage structure coupled with fractal analysis has shown that these shifts in growth behavior seem
to be correlated with changes in the damage morphology.
Laser damage resistance is a key factor for the improvement of high power laser system. The PETAL laser, developed by the CEA-CESTA (France), uses meter scale reflective optics to compress, transport and focalize sub-picosecond laser pulses at 1053nm with high-energy [1]. In the case of defect-free material, laser-induced damage in the sub-picosecond regime is known to be deterministic since the threshold depends only on the electronic structure of the irradiated materials, the pulse duration and the enhancement of the electric fields in thin film coatings. Based on this consideration, a mono-shot technique has been investigated to assess the intrinsic damage resistance of optical component with only one laser shot. On the other hand, while considering real optical components, manufacturing processes included nanoscale defects in the functional coating. These defects can be ejected when irradiated and strongly reduce the laser damage resistance of optics: rasterscan procedure has then been developed to determine defect-induced damage densities. These densities are found to be high even for fluences well below the intrinsic Laser-Induced Damage Threshold and they increase with the fluence. These experiments bring new information on the operating characteristics of optics in short pulse regime. Once damage is triggered, its evolution under subsequent irradiations has also been studied. Growth experiments have been compared to numerical simulations. The investigations on growth behavior allow a better estimation of the functional lifetime of an optic in its operating conditions. The whole of results, damage initiation and damage growth, is discussed to the light of the laser damage observed on PETAL optics.
The original damage ring pattern at the exit surface of fused silica induced by highly modulated nanosecond infrared laser pulses demonstrates the time dependence of damage morphology. Such a damage structure is used to study the dynamics of the plasma issued from open cracks. This pattern originates from electron avalanche in this plasma, which simultaneously leads to an ionization front displacement in air and a silica ablation process. Experiments have shown that the propagation speed of the detonation wave reaches about 20 km/s and scales as the cube root of the laser intensity, in good agreement with theoretical hydrodynamics modeling. During this presentation, we present the different phases and the associated mechanisms leading to this peculiar morphology: • During an incubation phase, a precursor defect provides energy deposit that drives the near surface material into a plasma state. • Next the silica plasma provides free electrons in the surrounding air, under laser irradiation an electron avalanche is initiated and generates a breakdown wave. • Then this breakdown wave leads to an expansion of the air plasma. This latter is able to heat strongly the silica surface as well as generate free electrons in its conduction band. Hence, the silica becomes activated along the breakdown wave. • When the silica has become absorbent, an ablation mechanism of silica occurs, simultaneously with the air plasma expansion, resulting in the formation of the ring patterns in the case of these modulated laser pulses. These mechanisms are supported by experiments realized in vacuum environment. A model describing the expansion of the heated area by thermal conduction due to plasma free electrons is then presented. Next, the paper deals with the two damage formation phases that are distinguished. The first phase corresponds to the incubation of the laser flux by a subsurface defect until the damage occurrence: an incubation fluence corresponds to this phase. The second is related to the damage expansion that only refers to the energy deposit feeding the activation mechanism up to the end of the pulse: an expansion fluence corresponds to this phase. A striking feature is that the damage diameters are proportional to the fluence of expansion at a given shot fluence. Indirectly, the fluences of incubation by the precursors are then determined.
In the context of high power laser systems, the laser damage resistance of fused silica surfaces at 351 nm in the nanosecond regime is a major concern. Under successive nanosecond laser irradiations, an initiated damage can grow which can make the component unsuitable. The localized CO2 laser processing has demonstrated its ability to mitigate (stopping) laser damage growth. In order to mitigate large damage sites (millimetric), a method based on fast microablation of silica has been proposed by Bass et al. [Bass et al., Proc. SPIE 7842, 784220 (2010)]. This is accomplished by scanning of the CO2 laser spot with a fast galvanometer beam scanner to form a crater with a typical conical shape. The objective of the present work is to develop a similar fast micro-ablation process for application to the Laser MegaJoule optical components. We present in this paper the developed experimental system and process. We describe also the characterization tools used in this study for shape measurements which are critical for the application. Experimental and numerical studies of the downstream intensifications, resulting of cone formation on the fused silica surface, are presented. The experimental results are compared to numerical simulations for different crater shape in order to find optimal process conditions to minimize the intensifications in the LMJ configuration. We show the laser damage test experimental conditions and procedures to evaluate the laser damage resistance of the mitigated sites and discuss the efficiency of the process for our application.
Based on an experimental system that can be used for simultaneous laser damage testing and time-resolved acquisition of intensity and phase images, we describe different experiments related to the study of laser damage process in the sub-picosecond regime. We report firstly on quantitative measurement of the Kerr effect in a fused silica substrate at fluences closed to the Laser Induced Damage Threshold. Then we study the damage initiation process in optical coatings, linked to intrinsic properties of the materials, and the dynamics of free electron generation and relaxation. At last, damage growth sequences are analyzed with time-resolved microscopy in order to understand laser damage growth in the case of High Reflective mirrors.
We report on the development of a mitigation process to prevent the growth of UV nanosecond laser-initiated damage sites under successive irradiations of fused silica components. The developed process is based on fast microablation of silica as it has been proposed by Bass et al. [Bass et al., Proc. SPIE 7842, 784220 (2010)]. This is accomplished by the displacement of the CO2 laser spot with a fast galvanometer beam scanner to form a crater with a typical conical shape to mitigate large (millimetric) and deep (few hundred microns) damage sites. We present the developed experimental system and process for this application. Particularly, we detail and evaluate a method based on quantitative phase imaging to obtain fast and accurate three-dimensional topographies of the craters. The morphologies obtained through different processes are then studied. Mitigation of submillimetric nanosecond damage sites is demonstrated through different examples. Experimental and numerical studies of the downstream intensifications, resulting in cone formation on the surface, are presented to evaluate and minimize the downstream intensifications. Eventually, the laser damage test resistance of the mitigated sites is evaluated at 355, 2.5 ns, and we discuss on the efficiency of the process for our application.
We designed an optical coating based on TiO2/SiO2 layer pairs for broad bandwidth high reflection (BBHR) at 45-deg angle of incidence (AOI), P polarization of femtosecond (fs) laser pulses of 900-nm center wavelength, and produced the coatings in Sandia’s large optics coater by reactive, ion-assisted e-beam evaporation. This paper reports on laser-induced damage threshold (LIDT) tests of these coatings. The broad HR bands of BBHR coatings pose challenges to LIDT tests. An ideal test would be in a vacuum environment appropriate to a high energy, fs-pulse, petawatt-class laser, with pulses identical to its fs pulses. Short of this would be tests over portions of the HR band using nanosecond or sub-picosecond pulses produced by tunable lasers. Such tests could, e.g., sample 10-nm-wide wavelength intervals with center wavelengths tunable over the broad HR band. Alternatively, the coating’s HR band could be adjusted by means of wavelength shifts due to changing the AOI of the LIDT tests or due to the coating absorbing moisture under ambient conditions. We had LIDT tests performed on the BBHR coatings at selected AOIs to gain insight into their laser damage properties and analyze how the results of the different LIDT tests compare.
Some silica plates of high power nanosecond lasers may be a few centimeter thick for instance because they should sustain vacuum. Measuring laser-induced damage thresholds at the output surface of these thick silica plates is a complex task because non-linear laser propagation effects may occur inside the plate which prevents knowing accurately the fluence at the output. Two non-linear effects have to be considered: stimulated Brillouin scattering (SBS) and Kerr effect. SBS is mainly driven by the spectral power density of the pulses: if the spectral power density is below a threshold, SBS is negligible. Thus, spectral broadening is required. Kerr effect depends on the instantaneous intensity. Hence, a smooth temporal shape without overshoots is required. However, both conditions (wide spectrum and no overshoots) are impossible to fulfill with standard lasers. As a matter of fact, an injected laser has a smooth temporal profile but is spectrally narrow. Without injection, the laser is multimode yielding a wide spectrum but a chaotic temporal profile. We solved the problem by phase-modulating a continuous-wave seeder of our laser (patent pending). The phasemodulation frequency is adjusted to a multiple of the inverse of the round-trip time of the laser cavity. The laser pulses have a wide spectrum to suppress SBS and do not exhibit temporal overshoots to reduce Kerr effects. During the presentation, we will show the features of the laser pulses and laser-induced damage measurements of thick silica plates using this scheme.
Seeded nanosecond Q-switched Nd:YAG lasers working with an unstable resonator and a variable-reflectivity-mirror are widely used for they represent useful sources for stable and repeatable light-matter-interaction experiments. Moreover, in most setups, the fundamental wavelength is converted to higher harmonics. When the injection seeder is turned off, random longitudinal mode beating occurs in the cavity, resulting in strong variations of the temporal profile of the pulses. The generated spikes can then be ten times higher than the maximum of equivalent seeded pulses. This strong temporal incoherence is shown to engender spatial incoherence in the focal plane of such unseeded pulses leading to an instantaneous angular displacement of tens of µrad. This effect is even more pronounced after frequency conversion.
The influence of vacuum on nanosecond laser-induced damage at the exit surface of fused silica components is investigated at 1064 nm. In the present study, as previously observed in air, ring patterns surrounding laserinduced damage sites are systematically observed on a plane surface when initiated by multiple longitudinal modes laser pulses. Compared to air, the printed pattern is clearly more concentrated. The obtained correlation between the damage morphology and the temporal structure of the pulses suggests a laser-driven ablation mechanism resulting in a thorough imprint of energy deposit. The ablation process is assumed to be subsequent to an activation of the surface by hot electrons related to the diffusive expansion of a plasma formed from silica. This interpretation is strongly reinforced with additional experiments performed on an optical grating in vacuum on which damage sites do not show any ring pattern. Qualitatively, in vacuum, the intensity-dependent ring appearance speed V ∝ I1/2 is shown to be different than in air where V ∝ I1/3 . This demonstrates that the mechanisms of formation of ring patterns are different in vacuum than in air. Moreover, the mechanism responsible of the propagation of the activation front in vacuum is shown to be outdone when experiments are performed in air.
A rasterscan procedure is set to determine laser-induced damage densities in sub-picosecond regime at 1053nm on high-reflective coatings. Whereas laser-induced damage is usually considered deterministic in this regime, damage events occur on these structures for fluences lower than their intrinsic Laser-Induced Damage Threshold (LIDT). Damage densities are found to be high even for fluences as low as 20% of the LIDT. Scanning Electron Microscope observations of these “under threshold” damage sites evidence ejections of defects, embedded in the dielectric stack. It brings a new viewpoint for the qualification of optical components and for the optimization of manufacturing processes of coatings.
We have designed and produced an optical coating suitable for broad bandwidth high reflection (BBHR) at 45° angle of
incidence (AOI), P polarization (Ppol) of petawatt (PW) class fs laser pulses of ~ 900 nm center wavelength. We have
produced such BBHR coatings consisting of TiO2/SiO2 layer pairs deposited by ion assisted e-beam evaporation using
the large optics coater at Sandia National Laboratories. This paper focuses on laser-induced damage threshold (LIDT)
tests of these coatings. LIDT is difficult to measure for such coatings due to the broad range of wavelengths over which
they can operate. An ideal test would be in the vacuum environment of the fs-pulse PW use laser using fs pulses
identical to of the PW laser. Short of this ideal testing would be tests over portions of the HR band of the BBHR coating
using ns or sub-ps pulses produced by tunable lasers. Such tests could be over ~ 10 nm wide wavelength intervals whose
center wavelengths could be tuned over the BBHR coating’s operational band. Alternatively, the HR band of the BBHR
coating could be adjusted by means of wavelength shifts due to changing the AOI of the LIDT tests or due to absorbed
moisture by the coating under ambient conditions. We conduct LIDT tests on the BBHR coatings at selected AOIs to
gain insight into the coatings’ laser damage properties, and analyze how the results of the different LIDT tests compare.
While considering long pulse or short pulse high power laser facilities, optical components performances and in particular laser damage resistance are always factors limiting the overall system performances. Consequently, getting a detailed knowledge of the behavior of these optical components under irradiations with large beam in short pulse range is of major importance. In this context, a Laser Induced Damage Threshold test facility called DERIC has been developed at the Commissariat à l’Energie Atomique et aux Energies Alternatives, Bordeaux. It uses an Amplitude Systemes laser source which delivers Gaussian pulses of 500 fs at 1053 nm. 1-on-1, S-on-1 and RasterScan test procedures are implemented to study the behavior of monolayer and multilayer dielectric coatings.
The rasterscan procedure, developed to test large components, is an efficient method that allows measuring extremely low surface damage density (until 0.01 site/cm2 for large optics). This procedure was improved in terms of accuracy. The equipment, test procedure and data analysis to perform this damage test of large aperture optics are described. The originality of the refined procedure is that a shot to shot correlation is performed between the damage occurrence and the corresponding fluence by recording beam parameters of hundreds of thousands of shots during the qualification. Because tests are realized with small Gaussian beams (about 1mm @ 1/e), beam overlap and beam shape are key parameters which have to be taken into account in order to determine damage density. After complete data analysis and treatment, a repeatable metrology has been reached. The measurement is destructive for the sample. However the consideration of error bars on defects distributions allows us to compare data obtained on a same batch of optical components. This will permit to reach reproducible metrology. Then this procedure provides a straightforward means of comparing the experimental results obtained from several facilities using different lasers. Recently, an additional step has been added to the procedure, a growth step that permits considering only growing damage sites. To the end the lifetime of large optics on high power laser can be predicted.
Laser induced damage of optical components is often a limiting factor for the development of high power lasers. Indeed,
for many years, organic contamination is identified as a factor decreasing the laser induced damage threshold of optical
surfaces, limiting the use of high fluencies. Also, for the development of its laser facilities, Laser MégaJoule and
PETawatt Aquitaine Laser, the Commissariat à l’Energie Atomique et aux Energies Alternatives investigates the
influence of organic contamination on the performances of the optical components. Actually, although great care is
provided on the cleanliness of the optics, organic volatile compounds outgassed from surrounding materials can be
adsorbed by the sensitive surfaces during its timelife. Thus, for this study, performances of clean and contaminated
multilayer dielectric mirrors are compared. Contamination is intentionally realized either by controlled protocols or by
exposing optics inside the laser facilities. Qualification and quantification of the organic contamination is realized by
automated thermal desorption and gas chromatography coupled with mass spectrometry. Laser induced damage
threshold of clean and contaminated mirrors are then investigated by 1053 nm laser at 670 fs.
The morphology of laser-induced damage sites at the exit surface of fused silica is tightly correlated to the mode
composition of the nanosecond laser pulses at 1064 nm. In the single longitudinal mode (SLM) configuration, a molten
and fractured central zone is surrounded by a funnel-shaped surface modification. Ring patterns surround the damage
sites when these are initiated by multiple longitudinal modes (MLM) laser pulses. In this last mode configuration, the
pulses temporal profiles as well as the damage ring patterns differ from pulse to pulse. The appearance chronology of the
rings is found to be closely related to the temporal shape of the laser pulses. This supports that the damage morphology
originates from the coupling of a laser-supported detonation wave propagating in air with an ablation mechanism in
silica. In our experiments, the propagation speed of the detonation wave reaches about 20 km/s and scales as the cube
root of the laser intensity, in good agreement with theory.
Laser damage growth on the exit surface of fused silica optics is considered as exponential, the growth coefficient depending essentially on fluence. In this presentation, experiments with large beams have been carried out at 351 nm under nanosecond pulses. A statistical analysis has then been conducted leading to a refined representation of the growth. The effect of several parameters has also been taken into account to describe precisely the growth phenomenon. Finally, the two main parameters reporting the growth are the mean fluence and the size of the damage sites. Contributions of other parameters have been estimated too: the number of neighbors around the damage site, the shot number.... From experimental results, a model based on a statistical approach has been developed that permits the description of a complete sequence of growth. At the end, the knowledge of damage initiation and damage growth permit the determination of the lifetime of optical components illuminated with successive shots.
With the purpose of understanding nanosecond laser induced damage mechanisms when working with multiple longitudinal mode pulses, an accurate measurement of the temporal profiles is required. In this study, the use of a streak camera with a wide bandwidth is justified through the knowledge of the Nd:YAG spectral characteristics. A statistical and phenomenological analysis of multiple longitudinal modes intensity profiles is then performed through experiments and modeling. The resolution limitation of our photodiodes is also discussed.
The laser induced damage densities measured on fused silica surface are found to be higher when produced with multiple longitudinal mode pulses than those produced by single longitudinal mode pulses at 1064 nm. The enhancement of the three-photon absorption due to the intensity spikes related to longitudinal mode beating might favor the damaging process at this wavelength. At 355 nm the picture is different. The absorption is supposed to be linear and an opposite behavior occurs. Evidences of a process leading to the possible annealing of a part of absorbent defects are discussed in this paper.
The determination of surface damage densities of thick optical components is tricky due to the
occurrence of non-linear effects (Brillouin and Kerr) that affect the beam propagation through the
optics. It is then compulsory to record the beam parameters, mainly the temporal profile, in order to
predict and calculate fluence and/or intensity on the rear surface taking into account the non-linear
beam propagation.
Experiments have been realised with the use of large beams and several phase modulations were
activated, leading to numerous peak intensities due to the occurrence of temporal amplitude
modulations. Results are first compared in the case of thin optics in order to separate the intrinsic
absorptions by the defects which are the weak points of the optics to the effect of the non-linear
propagation. The correspondence between the length of the filaments and the beam parameters has been
realised in order to highlight the relevant beam parameters that have to be considered for the damage
test of thick optics. The whole of measurements and modeling permit us to measure more accurately the rear surface
damage of thick optics due to intrinsic defects.
The Laser Mégajoule (LMJ) facility has about 40 large optics per beam. For 22 bundles with 8 beams per bundle, it will contain about 7.000 optical components. First experiments are scheduled at the end of 2014. LMJ components are now being delivered. Therefore, a set of acceptance criteria is needed when the optical components are exceeding the specifications. This set of rules is critical even for a small non-conformance ratio. This paper emphasizes the methodology applied to check or re-evaluate the wavefront requirements of LMJ large optics. First we remind how LMJ large component optical specifications are expressed and we describe their corresponding impacts on the laser chain. Depending on the location of the component in the laser chain, we explain the criteria on the laser performance considered in our impact analyses. Then, we give a review of the studied propagation issues. The performance analyses are mainly based on numerical simulations with Miró propagation simulation software. Analytical representations for the wavefront allow to study the propagation downstream local surface or bulk defects and also the propagation of a residual periodic aberration along the laser chain. Generation of random phase maps is also used a lot to study the propagation of component wavefront/surface errors, either with uniform distribution and controlled rms value on specific spatial bands, or following a specific wavefront/surface Power Spectral Distribution (PSD).
Various scratches on fused silica optics after polishing have been characterized with confocal microscopy and then tested
with nanosecond UV laser. Scratches are identified as a major contributor to laser damage even if they are only a few
micrometers wide. We propose a process in order to remove these scratches whose depth ranges from 2 to 16 μm. We
use a CO2 laser to heat the scratched areas at high temperature which will heal fractures due to viscous flow. The
characterizations were completed by laser damage tests that finally prove the effectiveness of the repair. We conclude
also that this repair process proves to be fast, localized to the scratch and clean because no debris are generated.
The comparison of laser-damage-densities (LDD) measurements performed with pulsed laser radiation at different facilities is tricky due to numerous parameters involved. These parameters, namely pulse length, profile, and frequency, beam size, as well as method of damage detection have significant impact on final result. Previous methods and suitable data processing developed to determine with accuracy and repeatability the LDD allow us to achieve this comparison, e.g. the reproducibility. Since such studies are related to the life-time predictions for large aperture optical materials used in high-power lasers, the question that is addressed in this presentation concerns the representativeness of such results as regards of laser damage with large and real beams. Tests with large beams of centimetric size on a high power laser facility have beam performed according to a parametric study and are compared to small beam laboratory tests. More the emphasis is on the optical component thickness that may affect both damage initiation and damage growth due to the occurrence of non-linear effects that intensify damage issue.
KEYWORDS: Laser induced damage, Crystals, Resistance, High power lasers, Second-harmonic generation, Laser crystals, Frequency converters, Bulk lasers, Diffraction, Frequency conversion
We have laser conditioned a couple of KDP-SHG and DKDP-THG samples thanks to a facility which delivers 6 ns
fundamental (1,053 nm, noted 1ω) pulses, and the harmonics generated by the crystals themselves. The conditioning
ramp has been established according to a model coupling statistics and heat transfer, in order to minimize the generation
of bulk laser damage during the process. Then the efficiency of this procedure has been evaluated for both samples using
two laser damage testing setups, and compared to the best monochromatic conditioning process known to date. For the
KDP-SHG, it appears that this procedure is less efficient than the monochromatic conditioning. But it raises the
resistance to laser damage of the SHG to a level compatible with the use on megajoule-class high power lasers. For the
DKDP-THG, the efficiency of both procedures is quite similar. And even if the conditioned THG still exhibits laser
damage within the range of high power laser working fluences at 351 nm, the density is only a few per mm3.
Nanosecond laser-induced damage (LID) in potassium dihydrogen phosphate (KH2PO4 or KDP) remains an issue for
light-frequency converters in large-aperture lasers such as NIF (National Ignition Facility, in USA) and LMJ (Laser
MegaJoule, in France). In the final optic assembly, converters are simultaneously illuminated by multiple wavelengths
during the frequency conversion. In this configuration, the damage resistance of the KDP crystals becomes a crucial
problem and has to be improved. In this study, we propose a refined investigation about the LID mechanisms involved in
the case of a multiple wavelengths combination. Experiments based on an original pump-pump set-up have been carried
out in the nanosecond regime on a KDP crystal. In particular, the impact of a simultaneous mixing of 355 nm and 1064
nm pulses has been experimentally studied and compared to a model based on heat transfer, the Mie theory and a Drude
model. This study sheds light on the physical processes implied in the KDP laser damage. In particular, a three-photon
ionization mechanism is shown to be responsible for laser damage in KDP.
Previous work on KDP has shown that thermal annealing could improve laser damage resistance of KDP optics at 3w. However, the improvement varies with the pulse length: whereas a strong improvement was observed at 16ns, no improvement at all was observed for a pulse length of 2.5ns. Whatever the pulse length, though, combinations of laser conditioning and thermal annealing led to better results than laser conditioning alone. The goal of this study is to verify if these results also hold for DKDP. A major difference is that, due to quadratic to monoclinic high temperature transition, the annealing temperature considered for KDP cannot be applied to KDP. This paper reports the temperature range considered for DKDP as well the modifications brought by thermal annealing on laser damage resistance at 12ns and 2.5ns.
Nanosecond Laser-Induced Damage (LID) in potassium dihydrogen phosphate (KH2PO4 or KDP) remains an
issue for light-frequency converters in large-aperture lasers such as NIF (National Ignition Facility, in USA) and
LMJ (Laser MegaJoule, in France). In the final optic assembly, converters are simultaneously illuminated by
multiple wavelengths during the frequency conversion. In this configuration, the damage resistance of the KDP
crystals becomes a crucial problem and has to be improved. In this study, we propose a refined investigation
about the LID mechanisms involved in the case of a multiple wavelengths combination. Experiments based on an
original pump-pump set-up have been carried out in the nanosecond regime on a KDP crystal. In particular, the
impact of a simultaneous mixing of 355 nm and 1064 nm pulses has been experimentally studied and compared
to a model based on heat transfer, the Mie theory and a Drude model. This study sheds light on the physical
processes implied in the KDP laser damage. In particular, a three-photon ionization mechanism is shown to be
responsible for laser damage in KDP.
We investigate the influence of THG-cut KDP crystal orientation on laser damage at 1064 nm under nanosecond
pulses. This study makes a connection between precursor defects and the influence of their orientation on the
laser damage. Previous investigations have already been carried out in various crystals and particularly for KDP,
indicating propagation direction and polarization dependences. We performed experiments for two orthogonal
positions of the crystal and results clearly indicate that KDP crystal laser damage depends on its orientation.
We carried out further investigations on the effect of the polarization orientation, by rotating the crystal around
the propagation axis. We then obtained the evolution of the damage probability as a function of the rotation
angle. To account for these experimental results, we propose a model based on heat transfer, the Mie theory
and a Drude model. The geometry of the precursor defects is assumed to be ellipsoid-shaped and we numerically
introduce absorption efficiency calculations for this geometry. Modeling simulations are in good agreement with
experimental results.
CO2 laser is used to prolong the lifetime of large optics for high power lasers such as the NIF and LMJ.
Indeed, on silica optical components, damaged sites, whose diameter is in the order of tens of microns, appear at
high UV laser fluence, and the size of such sites increases exponentially with each UV laser shot. An intense
heat by CO2 laser ejects the material from the surface of the optical component and removes all fractures around
the damaged site so that this site will not be damaged at fluences of operation of the UV laser. A crater is formed
at the site of initial damage. But the intense heat creates debris and residual stress around this crater. Due to these
debris and stress, the optical component is again weakened. We show here that a second heating process, done
with different settings of the CO2, named here laser annealing, eliminates the debris and reduce stress. The
results presented here establish that annealing significantly improves the resistance of laser optics.
The rasterscan test procedure implemented in order to determine low laser damage density of large aperture UV
fused silica optics was improved in terms of accuracy and repeatability. Tests have been carried on several facilities
using several pulse durations and spatial distributions. We describe the equipment, test procedure and data analysis to
perform this damage test with small beams (Gaussian beams, about 1mm @ 1/e, and top hat beams). Then, beam overlap
and beam shape are the two key parameters which are taken into account in order to determine damage density. After
data analysis and treatment, a repeatable metrology has been obtained. Moreover, the consideration of error bars on
defects distributions permits to compare data between these installations. This allows us to look for reproducibility, a
necessary condition in order to share results and to make reliable predictions of laser damage resistance. For that, a
careful attention has been paid to beam analysis.
A rasterscan test procedure [L. Lamaignère et al, Rev. Sci. Instrumen. 78, 103105 (2007)] has been
implemented in order to determine low laser damage density of large aperture UV fused silica optics. This
procedure was improved in terms of accuracy and repeatability. Tests have been carried on several facilities
using several pulse durations and spatial distributions. We describe the equipment, test procedure and data
analysis to perform this damage test with small beams (Gaussian beams, about 1mm @ 1/e, and top hat beams).
Then, beam overlap and beam shape are the two key parameters which are taken into account in order to
determine damage density. After data analysis and treatment, a repeatable metrology has been obtained.
Moreover, the consideration of error bars on defects distributions permits to compare data between these
installations. This allows us to reach reproducibility, a necessary condition in order to share results and to make
reliable predictions of laser damage resistance.
Laser damage in KDP crystals has been studied since several years and more accurately with emergence of projects
like LMJ (Laser MégaJoule, in France) or NIF (National Ignition Facility, in US). Laser damage tests are essentially
performed at 351-nm wavelength (3ω), with regards to their optical behaviours on forementioned facilities. But only few
data are available at 1064 nm (1ω) and at 532 nm (2ω), and even with wavelength-mixing more representative of
operational conditions of KDP crystals. So in a first approach, we tried to carry out an identity chart of the crystal by
performing mono-wavelength tests at 1ω, 2ω and 3ω. Then, a campaign of combination of multi-wavelength (typically
3ω and 1ω) tests has been started with several temporal delays between 3ω and 1ω pulses. These first results lead us to
improve pre-existing modelling codes developed by CEA, which have proved their robustness to 3ω -experiment results.
Foremost interests consist in implementing wavelength dependency and energy deposition mechanism as a consequence
of our first observations on KDP.
During the development of the laser megajoule (LMJ), a high power laser facility dedicated to DT fusion, CEA has made
important efforts to understand and improve laser induced damage threshold of fused silica optics at the wavelength of
351 nm. For several years, with various industrials and academics partners, we have focused on optimizing the grinding,
lapping and polishing processes to increase materials performance. In this paper, we describe our efforts in various
fields: subsurface damage characterization, lapping process simulation, diamond grinding and lapping machine
instrumentations, ... Our concern is to control and manage the material removal at each step of the process in order to
reduce the cracks region extension and thus to diminish the damage density.
The lifetime of silica optics in high power laser facility as the Laser MégaJoule (LMJ) is typically limited by the
initiation of surface damages and their subsequent growth. To prevent this problem, a mitigation technique is used: it
consists in a local melting of silica by CO2 laser irradiation on the damage site. Because of the difficulty to produce
efficient mitigated sites with large depth, the characterization of damage site to mitigate is very important. In this
context, confocal microscopy appears to be an efficient solution to detect precisely cracks present under the damage site.
In order to characterize the effect of thermal annealing on laser damage resistance of KDP,
several combinations of laser conditioning and thermal annealing were applied to two SHG KDP
samples. One sample was tested at 3ω, 16ns and the other one at 3ω, 2.5ns. Results show that
whereas thermal annealing improves laser damage for a 16ns pulse, no effect can be measured at a
pulse length of 2.5ns. Combining laser conditioning and thermal annealing has a stronger effect
on laser damage resistance than laser conditioning alone, even for a 2.5ns pulse length for which
thermal annealing was found to have little or no influence. It was also found that for a short pulse
length maximum gain was obtained when thermal annealing was applied after laser conditioning.
A rasterscan test procedure [L. Lamaignère et al, Rev. Sci. Instrumen. 78, 103105 (2007)] has been implemented
in order to determine low laser damage density of large aperture UV fused silica optics. This procedure was improved in
terms of accuracy and repeatability and is now used for the determination of bulk damage density for KDP crystals. The
large area (volume) scanned during tests permits to measure very low damage density. On small samples, small area are
tested using the normalized 1/1 test procedure consisting on the irradiation of few sites at several fluences. The classical
damage probability plot is converted in terms of damage density. The two testing procedures are complementary: the 1/1
mode is practical to test a wide fluence range while the rasterscan mode allows exploring low damage densities with
higher accuracy.
Tests have been carried out on several facilities using several pulse durations and spatial distributions. We
describe the equipment, test procedure and data analysis to perform this damage test with small beams (Gaussian beams,
about 1mm @ 1/e, and top hat beams). Then, beam overlap and beam shape are the two key parameters which are taken
into account in order to determine damage density. After data analysis and treatment, a repeatable metrology has been
obtained. Moreover, the consideration of error bars on defects distributions permits to compare data between these
installations. This allows us to reach reproducibility, a necessary condition in order to share results and to make reliable
predictions of laser damage resistance.
Other tests are realized with larger beams (centimeter sized) and with a single shot. Due to a large beam contrast,
a large fluence range is then covered. Then after data treatment, we find a good correlation between tests realised with
small and large beams. This allows us to make tests with different laser characteristics (spectral modulations, pulse
duration, laser polarisation) and then to study their influences on laser damage.
Significant improvement in polishing processes of fused silica optical components, has increased optics lifetime at the
wavelength of 351 nm. Nonetheless, for large laser operation facilities like the Laser MegaJoule (LMJ), zero defect
optics are not yet available. Therefore a damage mitigation technique has been developed to prevent the growth of
initiated damage sites: this technique consists in a local melting and evaporation of silica by CO2 laser irradiation on the
damage site. Because of the difficulty to produce efficient mitigated sites with large depth, the initial depth of damage to
mitigate is a critical issue. An aim of our work was to determine the real extension of the damage site (including
fractures) for different laser pulse durations between 3 ns and 16 ns and at different laser fluences. The fractures are nondetectable
in conventional microscopy. The depth of the damage can thus be underestimated. Hence confocal microscopy, was used to observe these sub-surface fractures and to measure precisely the depth of damage. Results show that the damage is 2 to 4 times wider than deeper and this ratio is independent of the pulse duration and of the fluence. With this new information, the mitigation process can now be optimized.
In this paper, we present various laser conditioning experiments which have been performed with KDP SHG and DKDP THG samples. The different conditioning facilities used delivered laser pulses at 351 nm in the nanosecond (from 3 to 12 ns) or in the sub-ns (600 ps) regime. Finally, the efficiency of the various conditioning protocols was compared: 526 nm-6 ns and 351 nm-3 ns damage tests were performed respectively on SHG and THG samples. The results show that laser-conditioning SHG KDP samples at 351 nm either with ns or sub-ns pulses allows reducing the laser damage density so that it becomes consistent with the specification of high power lasers. They also confirm that conditioning THG DKDP samples at 351 nm using sub-ns pulses is more efficient than using ns pulses.
Campaigns of laser damage tests at 1w of Nd-YAG laser (1064 nm), 3w and with a combination of these two
wavelengths, were conducted to complete previous existing data on damage growth in fused silica output surface.
It is known that UV light is very effective in inducing preexisting damage craters to grow. When both
wavelengths are present, the effect of 1w beam on damage growth depends on the delay between the IR and the UV
beam. When the 1w reaches the sample before the 3w, it has nearly no consequence on growth rate. On the opposite,
when the IR beam is delayed and strikes the sample after the 3w pulse, its energy simply adds to the UV in enhancing
damage growth.
Damage initiation is much more affected by 3w than 1w pulses. However, the number of surface damage craters
also increased by the addition of 1w photons to the UV beam.
As laser conditioning ever increases the performance of KDP optics with respect to laser induced bulk damage so that it can meet high-power laser specifications, it is expected that surface damage may become the next threat that must be dealt with. This paper presents new data on surface damage initiation and growth at 3w. A surface damage mitigation process based on the ball-dimpling method is presented along with the first results on the behavior of this process with respect to laser irradiation.
During the life of a high-power laser chain, optical components may be damaged due to local high fluence levels in the
inhomogeneous beam. The origin of the laser damage can be impurities, surface defects or flaws and cracks resulting
from polishing, or it may be produced by self-focusing in the component. The aim of this study is to better understand
the correlation between a surface crack on a silica optical component and laser damage. To accomplish this, calibrated
indentations were made on silica samples. Observations of the sites were made with an optical microscope, and three
different morphologies were recognized. Then the zones containing the indentations were irradiated (single shot mode)
with a Nd Yag laser at 355 nm for various fluences. Subsequent observations of the sites were made with an optical
microscope, with the aim of correlating site morphology and laser-induced damage. Some sites were believed to have
undergone laser conditioning. They were further irradiated (raster scan mode) at high fluence, and some evidence for a
laser conditioning effect was obtained.
To evaluate the impact of particulate contamination in laser induced damage of optical material, an
experimental program is established. The first step consists in the Ligne d'Integration Laser (LIL) particle
contamination sampling. Carbonated cellophane tapes, antireflection coated and uncoated silica samples were
inserted in the LIL laser chain, in six different zones to collect particles. The second step is the pollution
characterization. Polluted cellophane tapes are analysed by Scanning Electron Microscopy and Energy
Dispersive Spectrometry. The density and the nature of particles collected in the Amplification Section are
found to be homogenous throughout this section. The pollution collected in the Frequency Conversion and
Focusing system is more complex. One of its features is a larger proportion of silica particles. The last step
consists in the silica samples irradiation. Antireflection coated and uncoated silica samples are examined by
optical microscopy, then irradiated at 1064 nm or 355 nm and examined again. No damage growing under
several irradiations is observed. We show a cleaning effect efficient for particles larger than 20 microns.
We report on laser-induced damage threshold (LIDT) and UV-laser excited defect formation measurements in large
aperture KDP crystals developed as doublers and triplers for mega-Joule laser. Measurements of LIDT were performed
according to the ISO 11254-2 standard for repetitive pulses with duration ~ 4 ns and repetition rate of 10 Hz. The results
for different laser wavelengths (1064, 532 and 355 nm) and polarizations are presented. The largest LIDT was observed
for 532 nm pulses and the 1064 nm wavelength had a strong dependence on laser polarization. The LIDT values at 532
nm and 355 nm also depended on the crystal cutting angle, which is different for doublers and triplers. A comparison of
LIDT with earlier reported crystal absorptance at different wavelengths is also performed.
The UV-laser induced defect formation was investigated by the means of pump-probe technique. The excitation was
performed with a single pulse of ns Nd:YAG laser (355 or 266 nm wavelength) and probing with another Nd:YVO4 laser
system (532 nm) operating at 1kHz. This gave us a temporal resolution of 1ms. The transient absorption of defect states
relaxed non-exponentially and fully disappeared in ~10 s. A comparison is made between crystal grown by distinct
growth methods and between different laser polarizations. An influence of laser conditioning on UV induced defect state
formation is also revealed.
In this paper, we present different procedures of laser conditioning realized on KDP doubler crystals. First, components
are treated either with an excimer laser (SOCRATE facility, 351 nm, 12 ns) or a Nd: YAG laser (MISTRAL facility,
355 nm, 7 ns). Then damage tests are performed at 2ω (532 nm - 5 ns BLANCO facility) and 3ω (355 nm - 2.5ns
LUTIN facility) in order to estimate the conditioning gain for these two wavelengths.
For the best procedures, results show that it is possible to increase laser damage threshold at 532 nm so that it becomes
compatible with the nominal specifications of the LMJ. Moreover, tests realized at 355 nm highlight also an
encouraging improvement for the laser conditioning of tripler crystals.
A major issue in high power lasers for fusion is laser-induced damage on optical components. Since damage is often initiated by a surface crack resulting from polishing, it is important to understand the physics involved in this phenomenon. In this study, calibrated surface scratches have been realized on two silica samples using an indenter-scratcher apparatus. A variety of scratches have been tested by applying different speeds and forces on the scratcher needle. Optical microscope observations show that the scratches made at highest speed create irregular dashed lines. In addition, we have observed, at intermediate speed, an evolution in time of the scratches due to local stresses in silica.
One of the samples was irradiated by a Nd:YAG laser beam at 355 nm with the scratches on the exit surface. Microscope observations were made before and after irradiation.
Strong dependence on the scratch speed was observed on the local laser damage. Again, temporal evolution of the damage has been observed.
We studied filamentation, front surface damage and rear surface damage at 1064 nm and 351 nm with nanosecond pulses on a fused silica optical window. With temporally single-mode pulses, self-focusing occurs together with front surface damage, which is attributed to a Stimulated Brillouin Back Scattering (SBS) wave. The use of temporally multi-mode pulses suppresses the occurrence of front surface damage, and increases self-focusing. With single-mode pulses, the observation of filaments seems coherent with standard Kerr self-focusing effect, and can be understood according to the numerical treatment by Marburger et al, using non linear index values measured in other experiments. However, when multi-mode pulses were used, filaments occurred for much smaller peak intensities, by about a factor of 2. In this case, the non linear index causing self-focusing appears to be twice bigger. This second case is relevant to the situation of vacuum windows in high power laser installations, where the spectrum of light is widened to get rid of SBS. We discuss the physical effects that could be causing the enhancement of self-focusing.
KDP crystals are currently used for frequency conversion and Pockels cells in large aperture laser systems such as the LMJ and NIF. These different functions are obtained by cutting the KDP crystals with different orientations. We show by measuring the LIDT with three different facilities, that the cut angle plays a key role in the damage mechanism. Consistently with the three measurement set-ups, we demonstrate that the doublers have a weaker LIDT value than the triplers. The z-cut KDP samples have a LIDT higher than both the doublers and the triplers. These results are analyzed in terms of probed volumes and pulse duration.
One of the major issues met in the operating of high power lasers concerns the cleanliness of laser components. In this context, in order to assess laser-induced damage in presence of metallic particulate contamination, we study the behaviour of aluminum on a silica substrate. Model samples containing calibrated aluminum square dots of 50 x 50 μ2 have been deposited by photolithography on a silica substrate. The sample was irradiated by a Nd:YAG laser at 1064 nm with different fluences and also different numbers of shots on each dot. Then the initial aluminum dot zone and the surrounding silica were analyzed using Nomarski microscopy, profilometry and photothermal microscopy. Laser fluence is revealed to be a very important parameter for the behaviour of aluminum dots. For example, it is possible to find a fluence of irradiation where aluminum dots are blown off the substrate and only small modifications occur to silica. In this case, increasing the number of shots doesn't significantly affect the silica surface.
We report the measurements of the linear and non linear absorption at 1064, 532 and 355 nm in samples of KDP crystals fabricated with the rapid growth process developed for NIF and LMJ high power lasers. Measurements were performed according to the ISO11551 standard by the "pulse" or "gradient" calorimetric method using a pulsed, diodepumped, Q-switched Nd:YAG laser. Time resolved spectroscopy method was used for the investigation of defects formation, responsible for the non linear absorption at 355 nm.
We report the measurements of the linear absorption at 1064, 532, and 355 nm and non linear absorption at 355 nm in samples of KDP crystals fabricated with the rapid growth process developed for NIF and LMJ high power lasers.
A new facility has been designed to enhance laser damage resistance at 351 nm of large scale 3w KDP and silica optics by laser treatment. This facility is a prototype, and the process will be industrialized as a means of fabrication of the LMJ optics. The first step of the process is a conditioning/initiation step, which consists of a UV laser raster scan of the whole optics; the second step is a step of detection and analysis of damage possibly initiated during the previous step; the third step is a mitigation step, which consists of a local melting of the detected damage on silica surface, in order to stop their growth. The facility is equipped with a 3w Nd:YAG laser allowing the process of both KDP and silica. A CO2 laser is used for damage mitigation. Both the lifetime increase and the reduction of the process duration of large scale optics have been taken into account with a view to industrialize the process.
The occurrence of filaments in fused silica irradiated by UV laser light is well described by the product of light Intensity by Length of propagation in the material. For a spatially gaussian peak, in the well-known treatment by Marburger et al, this product is predicted to depend upon input power and non linear index. At a wavelength of 355 or 351 nm, the compilation of our past and present measurements give a smaller critical intensity by length product, i.e. a higher non linear index, than previously measured.
These values of non linear parameters allow for the prediction of rear surface damage on thick windows. The predictions compare well with damage probability measurements. Even when the intensity is not high enough to generate filaments, self-focusing is still the cause of damage, due to the increase of output intensity and fluence.
This paper presents data reduction on an experimental set-up that we have recently developed at CESTA, France. It has been implemented to analyze laser-induced damage on optics dedicated to the Megajoule laser project. Our goal is to measure the damage fluence on samples under tests, using a statistical approach on a very large number of sites.
The laser-induced damage density is accurately plotted as a function of laser fluence, by measuring the fluence of every single shot during the scan. This improvement of accuracy modifies dramatically the interpretation of the results that can be made, compared to raw data, considering the mean fluence only.
We perform thne conditioning of various KDP crystals with a XeF excimer laser working at 351 nm. We determine the maximum available excimer laser fluence for conditioning without damage initiation within the crystal. We demonstrate enhancement of the damage resistance with the increase of the cumulative excimer laser fluence. Using the conditioning parameters we show that the damage resistance is also dependent on the crystalline orientation of the KDP samples.
For high power laser applications like the "Laser Megajoule" facility under construction in France, laser-induced damage threshold (LIDT) in fused silica is a limitation. CEA has made efforts to improve LIDT at the wavelength of 351 nm. Polishing and post polishing processes have been optimized. Laser damage sites density was decreased by several orders of magnitude by combining different fabrication steps. In order to further enhance optical laser resistance and to remove damaged sites on full-size optics, several small-beam raster scanning techniques have been studied and developed to condition fused silica optics. To stop the growth of damage sites, a continuous CO2 laser was used to re-melt them. Laser induced damage tests, performed on instrumented and automated facilities, are reported in order to check and illustrate the effectiveness of these treatments. Damage initiation studies as well as damage growth measurements are presented.
Variable experimental conditions were used to measure the occurrence of front surface, rear surface and filamentation damage in synthetic fused silica windows. Experiments were performed at 355 nm with a table-top beam of mm-size, and at 351 nm with ALISE laser, a 100 J installation. The 351 nm beam was about 3 cm wide at the entrance surface; it was single-mode temporally, with or without a frequency modulation which has the function of widening the spectrum to decrease Stimulated Brillouin Scattering. The 355 nm was single-mode temporally. Thin windows showed very scarce front damage and no filament damage at intensities which cause a high density of rear surface damage. Without any spectral widening, the thicker windows (4.3 cm) showed appreciable amount of front surface damage; filaments were observed and but no filaments. When a spectral modulation was added, front surface damage vanished, filaments and rear surface damage were observed.
Experiments have been performed to measure the rate of laser-induced damage growth at the rear surface of fused silica windows at 1064, 1053 and 351 nm. One test bench delivered 9 ns monomode gaussian pulses at 10 Hz and 1064 nm. The size of the focused beam on the sample was a few mm2. Another test bench delivered 2.5 ns single or multimode pulses at 1053 and 351 nm. The focused spot on the sample was a few cm2. We compare and discuss our laboratory experimental results, the larger scale ALISE laser data and other results obtained at LULI.
The purpose of this paper is to gather experimental elements allowing for the prediction of laser damage on full size components installed on high power Nd-glass laser lines. Damage can initiated on material defects, which aren’t known in their nature, but the density of which can be measured. On transmissive optics, depending on the component thickness, and on the intensity distribution at the front surface, rear surface damage can also appear due to self-focusing of hot spots. These two contributions produce damage sites that are prone to grow. The growth rate has been shown to be proportional to the damaged area. The resulting exponential growth is the major limitation to the lifetime of optics. A representation of these phenomena in the plane Intensity/Fluence gives a practical description of the impact of laser damage on the lifetime of optical components. It also enlightens the comparison between different operating conditions.
The functional lifetime of large aperture optical components used in high power lasers, like LIL and LMJ facilities, is mainly determined by laser damage measurements. We present experimental procedures and statistical analysis, made on small samples with mm-size beams, to determine damage densities and damage growth laws. The tests and analysis are compared to other results obtained with larger beams (few cm2) on large aperture components.
In the field of the development of LIL and LMJ fusion class high power lasers, CEA has made important efforts to understand and improve laser induced damage threshold of fused silica optics at the wavelength of 351 nm. Since several years, we have focused on optimizing the grinding, polishing and post polishing processes to overcome the existing performances with various industrials and academics partners. In this paper, we describe our understanding of the nature of the polished silica interface and our approach to rich our damage threshold goal. Our efforts were mainly put on reducing the cracks region extension and removing or optimizing the polishing top layer. We give also some details on the influences of each of the polishing process from rough material grinding to post processing. We demonstrate that some order of magnitude in laser damage initiation density can be gained by combining appropriate fabrication steps.
The functional lifetime of large aperture components used in high power lasers, like LIL and LMJ facilities, is mainly determined by laser damage measurements. Automatic damage test benches allow to obtain more data in less time than traditional tests. We present, first experimental procedures and statistical analysis made on small samples with mm-size beams, to determine damage densities and damage growth laws. The presented methods are the usual 1on1, Non1, Ron1 and Son1 tests and more specially the raster scan procedure. The tests and analysis are compared to other results obtained with larger beams (few cm2) on large optics. We show that the exact knowledge of each shot parameters (energy, surface and pulse duration) permits to determine the damage growth rate (and then to predict the lifetime of each optics), to precisely study self-focusing phenomenon and more to finely observe pre-damage-levels. In this way, the main parameters like fluence or intensity are associated to the observed phenomenon. Moreover laser beam diagnostics, many diagnostics used for the detection and the observation of damage occurrence are equally very important. It is also necessary to develop test procedures entirely computed which permit to scan all the surface of a component and to acquire in real time the beam parameters and the results of laser-matter interaction. Experimental results are reported to illustrate what could be achieved on an instrumented and automated facility.
The growth of damage sites on the rear surface of fused silica plates was studied as a function of fluence and angle of incidence. At 1053 nm, a 70 J beam, 3 ns in pulselength, was directed to a 5 cm2 zone on a bare fused silica window. Initiation and growth was observed. The growth of previously initiated sites was also studied. Growth is exponential in nature. The experiments allow for the determination of the growth coefficient as a function of fluence. At 355 nm, damage sites were irradiated at various angles of incidence, with a tripled Nd:Yag laser, spatially Gaussian, 2.5 ns in pulselength. By fitting growth with an exponential law, it was determined that the relevant fluence for growth was that taken inside the material.
In order to exhibit the role of laser damage precursor centers in silica, gold particles ranging from 0.2mm to 0.8mm have been included between two coatings of SiO2 deposited on silica substrates. UV and IR nano-second pulsed irradiations on samples with different SiO2 over-layer thicknesses (2, 5, 10 mm) has been performed. The damage morphologies observed with Nomarski and atomic force microscopes have shown to be dependent on fluence, wavelength and SiO2 over-layer thickness. In addition a localized irradiation study using 6mm spot size allows to aim accurately on an isolated particle. The measurement is compared to the experiments presented for a few hundred micrometers spot size. Indeed in this case we have to consider that several particles are simultaneously irradiated. A comparison between the respective morphologies could inform us on potential collective effects of the particles. The choice of experimental test conditions, and the whole associated results will be presented and discussed.
Rear surface damage is known to occur preferentially to front surface damage on silica lenses and windows transmitting 1ω or 3ω light of a Nd:Yag or Nd:glass laser. Lab-scale Experiments were performed to assess the contribution of self-focusing to the measurement of damage statistics. The occurrence of filamentation in the bulk and the statistics of rear surface damage were studied at 1064 nm and 355 nm, for a pulse length around 3 ns. The measurements were performed on synthetic fused silica samples, 4 to 5 cm thick. The laser beam had dimensions around 0.6 mm at the sample. The results are interpreted with the help of the calculations made by J. H. Marburger et al. They are also compared with other experimental results.
Samples of deuterated and hydrogenated KDP were submitted to damage tests at 355 nm and 351 nm, with a Nd:Yag laser and the Xe-F line of a excimer laser. Bulk damage was observed; the statistical occurrence and the phenomenology of this type of damage was studied in various conditions. The crystals were raster-conditioned with the excimer source, which delivered a 16 ns pulselength. Laser-conditioning was performed with increasing fluences. The highest usable fluence was limited by the occurrence of surface damage on the front of the crystals. After this irradiation, a two-fold improvement of the damaging fluences was obtained when testing with the excimer beam. However, damage statistics were almost unchanged on the Nd:Yag installation, where the pulselength is about 3 ns. This result is discussed with respect to the large scale conditioning of crystals for high power lasers.
With the aim of observing and simulating laser initiated damage, ultra-pure silica "model" samples, seeded with gold nanoparticles (diameters 2 - 5 nm), were prepared and exposed to 3 ns laser pulses at wavelength 355 nm. These samples enable us to study the mechanism of damage initiation caused by inclusions of nanometric size, which is the typical size of defects occurring in optical quality glass. The samples were studied in a series of experiments using a time-of-flight mass spectrometer at Argonne National Laboratory. This installation is of great interest because it enables us to combine the laser irradiation of the sample with the chemical identification of material ejected from the exposed surface. An evaluation of the quantity of gold atoms emitted during irradiation can thus be obtained from the experimental results. These experimental data are completed with “Nomarski” and “atomic force” microscope observations, and then interpreted. In particular, a comparison is made to numerical simulations obtained with our DELPOR code. An encouraging result is the existence of a pre-damage phase at very low fluences that is not detectable by classical optical devices.
The validation of numerical simulations of laser induced damage of fused silica requires detailed knowledge of the different parameters involved in the interaction. To approach the problem, we have performed simulations of laser energy deposition in spherical metallic defects and the surrounding fused silica. Our code DELPOR takes into account various laser/defect induced absorption mechanisms of SiO2, such as radiative ionization, avalanche and multiphotonic ionization. We have studied crater formation produced by the absorber explosion with a 2-D Lagrange-Euler code taking into account crack formation and propagation in the brittle material. To validate our simulations, we have made and tested samples of ultra-pure silica thin film, containing gold nanoparticles of diameter 0.6 μm. The fused silica coating could have three different thickness. We compare experiment and simulations for two laser irradiations at wavelengths 0.351 and 1.053 μm.
Simulations of laser-fused silica interactions at 0.351 μm are a key issue in predicting and quantifying laser damage in large laser systems such as LIL and LMJ. Validation of numerical simulations requires detailed knowledge of the different parameters involved in the interaction. To concentrate on a simple situation, we have made and tested a thin film system based on calibrated gold nanoparticles (0.2-0.8 μm diameter) inserted between two silica layers. The fused silica overcoat was either 2 or 10 microns thick. We have performed simulations of laser energy deposition in the engineered defect (i.e. nanoparticle) and the surrounding fused silica taking into account various laser/defect induced absorption mechanisms of SiO2 (radiative ionization, avalanche and multiphotonic ionization). We have studied crater formation produced by the absorber explosion with a 2D Lagrange-Euler code taking into account crack formation and propagation in the brittle material. We discuss the influence of the defect depth (with respect to the surface) on the damage morphology. The simulations are compared with our experimental results.
High damage threshold hafnia/silica HR mirrors were damage tested. The transient reflectivity of these mirrors was studied during laser irradiation and particularly during catastrophic damage by top layer chipping. The tests were performed in R/1 mode on 50 sites in order to correctly assess the statistical behavior of the damage threshold. During the ramp, we observed the formation of a first plasma coupled to a modification of the reflected pulse without formation of any pit or chipping damage. For further irradiations at higher fluences, the optical properties of the mirror were unchanged up to the creation of a catastrophic chipping damage coming with a more intense plasma. The statistical distributions of fluences obtained in the two cases were different. For the moment, there is no proven correlation between first plasma and chipping: plasma detection cannot be used as a nondestructive quality control.
Simulations of laser matter interaction at intermediate energies is a key issue in predicting and quantifying laser damage in the LMJ or the NIF facilities. We have done simulations of laser interaction with several metals from the solid state to the plasma by solving the Helmholtz equation in our hydrodynamic code DELPOR. We are comparing our result to time dependent reflectivity measurements on Aluminum and Iron and discuss the influence of the transitions of phase of the materials in laser mater interaction.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.