Demonstration of stable, long-term operation of a nanosecond pulsed DPSSL at 10 J, 100 Hz

: We report on stable, long-term operation of a diode-pumped solid-state laser (DPSSL) amplifying 15 ns pulses at 1029.5 nm wavelength to 10 J energy at 100 Hz pulse rate, corresponding to 1 kW average power, with 25.4% optical-to-optical efficiency. The laser was operated at this level for over 45 minutes ( ∼ 3 · 10 5 shots) in two separate runs with a rms energy stability of 1%. The laser was also operated at 7 J, 100 Hz for 4 hours (1.44 · 10 6 shots) with a rms long-term energy stability of 1% and no need for user intervention. To the best of our knowledge, this is the first time that long-term reliable amplification of a kW-class high energy nanosecond pulsed DPSSL at 100 Hz has been demonstrated.


Introduction
Efficient, high energy (multi-J), high pulse rate (multi-Hz) nanosecond lasers are required for a wide range of applications.These include industrial materials processing and testing [1,2], development of inertial confinement fusion power plants [3,4] and high energy density physics research [5].High energy nanosecond lasers are also required as pump sources for high peak power (few 100s TW to PW) femtosecond amplifiers based on titanium-doped sapphire or optical parametric chirped-pulse amplification (OPCPA) technology [6].The resulting high peak power pulses in turn drive compact, high brightness radiation and particle sources [7][8][9] for imaging and medical applications [10][11][12].
High energy diode-pumped solid-state laser (DPSSL) technology has, over the last decade, reached the performance required for the implementation of these applications.Compared to traditional high energy flashlamp-pumped amplifier technology [13,14], DPSSLs offer significantly higher efficiency and pulse rates, thus increasing the parameter and application space of high energy laser technology.
In previous publications we presented the design and performance of lasers based on the Diode Pumped Optical Laser for Experiments (DiPOLE) concept, developed at the STFC's Central Laser Facility (UK) [15].DiPOLE is a laser amplifier architecture relying on diode-pumped, cryogenic gas cooled, multi-slab Yb:YAG gain medium.The DiPOLE concept was validated through the commissioning of a prototype laser generating 10 J, 10 ns pulses at 10 Hz pulse rate and its long-term stability and reliability were assessed by operating the laser at the 7 J, 10 Hz level for over 50 hours (achieved in separate runs, each lasting between 4 and 6 hours) [16].Energy scalability of the DiPOLE concept was demonstrated in two scaled-up DiPOLE100 lasers [17,18], delivering up to 150 J pulses at 10 Hz pulse rate [19].
The work reported in this paper focuses on pulse rate scaling of the DiPOLE concept from 10 Hz to 100 Hz and describes the design and performance of a DiPOLE laser amplifying nanosecond pulses to 10 J pulse energy at 100 Hz pulse rate.
The development of the DiPOLE-100Hz laser targets the needs of some industrial and scientific applications requiring high average power and high throughput rates.A tenfold increase in pulse rate brings about benefits in terms of higher processing rates and faster data acquisition.
A first demonstration of a DPSSL operating at the 10 J, 100 Hz level was achieved at the Institute of Laser Engineering, Osaka University in Japan; however, the laser was operated for only 20 seconds [20].
In this paper, we report on stable, long-term amplification of 15 ns pulses to 10 J pulse energy at 100 Hz, corresponding to 1 kW average power, at 1029.5 nm wavelength with 25.4% optical-to-optical efficiency.The system exhibited a long-term rms energy stability of 1% measured over 45 minutes (∼ 3 • 10 5 shots).To further assess long-term stability and reliability, the laser was operated at 7 J, 100 Hz for 4 hours, corresponding to 1.44 • 10 6 shots.No sign of laser-induced damage or component degradation was observed and the system did not require user intervention throughout the 4 hour test.These results demonstrate that DiPOLE technology is suitable for operation at 100 Hz pulse rate.To the best of our knowledge this is the first time that stable, long-term amplification of nanosecond pulses at 10 J, 100 Hz has been demonstrated.This work expands the parameter range of high energy DPSSLs, with expected benefits in a wide variety of fields, including manufacturing, materials testing and development of high peak power sources operating at high pulse rates.

DiPOLE-100Hz laser design
Figure 1(a) shows a schematic of the DiPOLE-100Hz laser chain showing typical output performance of each amplifier stage.Stages are grouped in two main sections: a front-end generating low energy pulses with precisely controllable temporal shape and a main cryogenic amplifier, based on the DiPOLE concept, to increase the pulse energy to 10 J.
The front-end seed source (Exail Modbox-FE-1030nm-40dB, France) relies on a continuouswave (CW) fiber oscillator generating a temperature-tuneable spectral output with a linewidth <10 MHz around 1029.5 nm.An acousto-optic modulator turns the CW emission into 300 ns duration pulses at 100 Hz pulse rate.The pulses are then amplified in fiber-amplifier stages before being temporally-shaped by a Mach-Zehnder electro-optic modulator controlled by an arbitrary waveform generator.The resulting output of the fiber seed source consists of few-nJ, arbitrarily shaped, pulses with duration between 1 and 100 ns at 100 Hz, delivered via a polarization-maintaining fiber.
A regenerative pre-amplifier (Körner LaserTechnologie, Germany), based on crystalline Yb:YAG gain medium pumped by pulsed 940 nm laser diodes, amplifies the seed pulses to around 4 mJ.The output 2 mm 1/e 2 diameter Gaussian beam is expanded and converted to a 6 mm super-Gaussian circular profile by a π-shaper (AdlOptica, Germany).The beam is then amplified to the 100 mJ pulse energy level in a room-temperature multi-pass diode-pumped Yb:YAG booster pre-amplifier.The 8 x 8 mm 2 output beam is magnified by a Keplerian telescope, propagated through a serrated aperture to obtain a 21.5 x 21.5 mm 2 square super-Gaussian profile, and delivered to the main cryogenic amplifier.The polarization of the beam is controlled by a quarter-and half-waveplate pair positioned before the serrated aperture.
Figure 2 shows a top-view schematic of the main amplifier architecture.The main amplifier uses 6 ceramic YAG slabs (Konoshima Chemical Co.Ldt., Japan) contained within an amplifier 920 nm and 1050 nm for angles of incidence between 0 • and 10 • .The slab set has a graded doping profile to ensure equalised pump energy absorption and uniform gain and thermal loading.
From the outside to the centre, the Yb-doping concentration levels are: 0.7 at-%, 1.0 at-% and 1.8 at-%.In each slab, the Yb-doping concentration is uniform throughout the entire Yb:YAG volume.The gain medium slabs are actively cooled by a stream of helium gas flown through 1.5 mm gaps between the slabs.The temperature of the gas is controlled by a cooling system (AF Cryo, France) containing a liquid nitrogen-based heat exchanger.
The gain medium slabs are held within a pressure vessel which forms part of the sealed helium circuit.Two 20 mm thick, dual-band AR (DBAR) coated, fused silica windows allow both the pump and the seed beams to propagate through the pressure vessel, as shown in the inset in Fig. 1b.The pressure vessel, like the entire helium circuit, is contained inside a vacuum vessel to provide thermal insulation.Optical access to the vacuum vessel is provided by two 12.7 mm thick DBAR-coated fused silica windows.
The gain medium slabs are face-pumped by two 939.5 nm laser diode sources operating at 100 Hz pulse rate (Amphos, Germany), each delivering up to 21 J pulse energy in 0.2 ms to 1 ms duration pulses.The pump beams are low divergence (5 • full angle) and are coupled into the amplifier head at a slight off-axis angle by high reflection mirrors operating at 40.5 • angle of incidence.At the centre of the amplifier head, the pump beam profile is a 24 x 24 mm 2 top hat.
The seed beam propagates 7 times though the amplifier head thanks to an angularly multiplexed geometry.On each pass, two Keplerian 1:1 telescopes (VSF1 and VSF2 in Fig. 2) using 900 mm head.Each slab is 5 mm thick with a 45 mm diameter Yb 3+ -doped YAG region surrounded by a 5 mm wide Cr 4+ -doped YAG absorptive cladding with an absorption coefficient of 6 ± 1 cm −1 at 1030 nm.The slabs are anti-reflection (AR) coated to achieve ≤ 0.2% reflectivity between 920 nm and 1050 nm for angles of incidence between 0 • and 10 • .The slab set has a graded doping profile to ensure equalized pump energy absorption and uniform gain and thermal loading.From the outside to the centre, the Yb-doping concentration levels are: 0.7 at-%, 1.0 at-% and 1.8 at-%.In each slab, the Yb-doping concentration is uniform throughout the entire Yb:YAG volume.The gain medium slabs are actively cooled by a stream of helium gas flown through 1.5 mm gaps between the slabs.The temperature of the gas is controlled by a cooling system (AF Cryo, France) containing a liquid nitrogen-based heat exchanger.
The gain medium slabs are held within a pressure vessel which forms part of the sealed helium circuit.Two 20 mm thick, dual-band AR (DBAR) coated, fused silica windows allow both the pump and the seed beams to propagate through the pressure vessel, as shown in the inset in Fig. 1(b).The pressure vessel, like the entire helium circuit, is contained inside a vacuum vessel to provide thermal insulation.Optical access to the vacuum vessel is provided by two 12.7 mm thick DBAR-coated fused silica windows.
The gain medium slabs are face-pumped by two 939.5 nm laser diode sources operating at 100 Hz pulse rate (Amphos, Germany), each delivering up to 21 J pulse energy in 0.2 ms to 1 ms duration pulses.The pump beams are low divergence (5 • full angle) and are coupled into the amplifier head at a slight off-axis angle by high reflection mirrors operating at 40.5 • angle of incidence.At the centre of the amplifier head, the pump beam profile is a 24 x 24 mm 2 top hat.Just before being absorbed within a water-filled beam dump, the output beam is sampled by collecting the leakage through a near-0 • high reflection mirror.The leaked beam is reduced in size and split by an uncoated wedged window.The beam transmitted through the wedged window is delivered to a calibrated energy meter.The reflected beam impinges onto a scatter disc (SM05CP2C, Thorlabs), with part of the scattered radiation collected by a fibre and delivered to a photodiode (DET08CFC, Thorlabs) for the characterisation of the temporal profile.

Beam polarisation uniformity diagnostics
The polarisation uniformity of the output beam was characterised following the methodology described in detail in [21].The polarisation purity was measured by propagating the seed beam from the front-end through the entirety of the main amplifier (7 passes through the amplifier head).The profile of the seed beam at the input to the main amplifier is acquired by a camera (Mako G-234, Allied Vision), further referred to as CAM1.The polarisation uniformity was The seed beam propagates 7 times though the amplifier head thanks to an angularly multiplexed geometry.On each pass, two Keplerian 1:1 telescopes (VSF1 and VSF2 in Fig. 2) using 900 mm focal length lenses perform relay-imaging of the beam onto the centre of the amplifier.To maintain good beam quality, weak spatial filtering is implemented on every pass by placing a tantalum plate with a 2 mm diameter aperture at the focal spot of the telescopes.Due to the high intensity of the beam at the focal spot, the tantalum plates are enclosed in vacuum to avoid breakdown of air.
An adaptive optic (AO) mirror (Dynamic Optics, Italy) positioned on the third pass compensates for static wavefront aberrations, measured after the seventh pass by a wavefront sensor (Phasics, France).
Near-field (NF) and far-field (FF) diagnostic cameras (Mako G-234, Allied Vision) monitor the beam at every pass.A high frame rate asynchronous dark-field (DF) diagnostic continuously monitors the condition of the amplifier head optics.
Just before being absorbed within a water-filled beam dump, the output beam is sampled by collecting the leakage through a near-0 • high reflection mirror.The leaked beam is reduced in size and split by an uncoated wedged window.The beam transmitted through the wedged window is delivered to a calibrated energy meter.The reflected beam impinges onto a scatter disc (SM05CP2C, Thorlabs), with part of the scattered radiation collected by a fiber and delivered to a photodiode (DET08CFC, Thorlabs) for the characterization of the temporal profile.

Beam polarization uniformity diagnostics
The polarization uniformity of the output beam was characterized following the methodology described in detail in [21].The polarization purity was measured by propagating the seed beam from the front-end through the entirety of the main amplifier (7 passes through the amplifier head).The profile of the seed beam at the input to the main amplifier is acquired by a camera (Mako G-234, Allied Vision), further referred to as CAM1.The polarization uniformity was characterized using the setup shown in the inset in Fig. 2, obtained by modifying the output NF and FF beam diagnostics.The leakage through a near-0 • high reflection mirror is reflected off a 45 • turning mirror (TM), with same reflectivity for both s-and p-polarization components, and propagated through a 1012 mm focal length lens (L1).L1 is followed by a quarter-and a half-waveplate (QWP and HWP), held in automated rotation mounts.Due to space constraints, the QWP and HWP could not be placed in the collimated beam.However, the long focal length of L1 ensures that the performance of QWP and HWP is not affected.The beam is re-collimated by a second lens (L2) and propagated through a polarising beam splitter cube (CM1-PBS253, Thorlabs), further referred to as analyser, held in an automated rotation mount and placed in front of the output NF diagnostic camera, referred to as CAM2.
Images from CAM1 and CAM2 were acquired using beam profiling software (RayCi, Cinogy) and analysed using a home-made LabVIEW algorithm.The algorithm calculates a normalized power P norm value as: where P CAM1 and P CAM2 are the integrated power measurements from CAM1 and CAM2, respectively.The calculation of P norm allows for compensation of power fluctuations in the input seed beam.The algorithm automatically adjusts the orientation of QWP and HWP to achieve maximum extinction (i.e.minimum P norm , referred to as P norm,min ).Then, the analyser is rotated by 90 • to achieve maximum transmission (i.e.maximum P norm , referred to as P norm,max ).The degree of polarization non-uniformity (DEP) of the output beam is calculated as: A perfectly uniform polarized beam is characterized by DEP = 0%, independent of the polarization state of the beam.In the worst case DEP = 50%, which is observed in a beam where the polarization state varies across the beam aperture, with equal portions of two orthogonal polarization states.

Results and discussion
When performing the experiments reported here, the room temperature multi-pass booster pre-amplifier (PA2 in Fig. 1) had not been commissioned.As a result, the seed energy available to the main amplifier was limited to the output of the regenerative pre-amplifier (PA1 in Fig. 1).After magnification via a Keplerian telescope and propagation through the 21.5 x 21.5 mm 2 serrated aperture, the beam from PA1 was directly delivered to the input of the main cryogenic amplifier.The polarization state of the beam at the input to the main cryogenic amplifier was chosen to minimize losses in the multipass.This was achieved by adjusting the quarter-and half-waveplates at the input of the main amplifier to maximize output pulse energy.PA1 was configured to amplify 15 ns pulses to 4.7 mJ at 100 Hz, resulting in 2.5 mJ pulses just after propagation through the beam transport optics and the serrated aperture.The beam profile at the input to the main amplifier is shown in Fig. 3.The asymmetry and the fringes characterising the seed beam profile were caused by alignment issues within PA1 which had not been rectified by the time the experiments were performed.
The Yb:YAG slabs within the main amplifier were cooled with helium gas at a temperature of 130 K, a mass flow rate of 200 g/s and a pressure of 14.5 bar(a).To maintain a constant thermal loading within the amplifier, the total pump energy and pump pulse duration were kept fixed at 39.4 J and 0.5 ms, respectively.The available total pump energy was lower than the expected 42 J due to the failure of one of the high current diode drivers.The effective pump energy was varied by controlling the relative delay between pump and seed pulses.Figure 4(a) shows the      measured output energy as a function of the effective total pump energy.An output energy of 10 J at 100 Hz was obtained with 39.4 J total effective pump energy, corresponding to 25.4% optical-to-optical efficiency, despite the reduced seed energy available.The laser was operated at this level for over 45 minutes in two separate runs, corresponding to a total of around 3 • 10 5 shots.The first run (after 23 minutes of operation) was interrupted because of laser-induced damage to one of the output 45 • mirrors, exposed to around 2 J/cm 2 beam fluence.After replacement with the same mirror type, no damage was observed during the second run.We believe that the damaged mirror was a defective component within the mirror batch.The data shows an average output energy of 9.93 J with rms stability of 1% (Fig. 4(b)).optical-to-optical efficiency, despite the reduced seed energy available.The laser was operated at this level for over 45 minutes in two separate runs, corresponding to a total of around 3 • 10 5 shots.
The first run (after 23 minutes of operation) was interrupted because of laser-induced damage to one of the output 45 • mirrors, exposed to around 2 J/cm 2 beam fluence.After replacement with the same mirror type, no damage was observed during the second run.We believe that the damaged mirror was a defective component within the mirror batch.The data shows an average output energy of 9.93 J with rms stability of 1% (Fig. 4b).Figure 5a shows the measured temporal pulse profile both at the input and at the output of the main amplifier during 10 J operation.During these experiments no attempt was made to optimise the seed temporal profile to achieve a flat-top output temporal profile.Figure 5b shows the output NF measurement at 10 J.As a result of gain saturation in the main amplifier, the output NF cross-section shows a relatively flat-top intensity distribution despite the non-uniformity in the input beam.The lower intensity at the corners of the output NF is due to the very low intensity of the same regions of the input beam profile.Figure 5c shows the output FF profile recorded without optimisation of the AO mirror.The full-width at half maximum (FWHM) angular spread of the central maximum in the FF image is 137  and 102  along the x-axis and the y-axis, respectively.This corresponds to 3.2 and 2.4 times the diffraction limit for a square 21.5 x 21.5 mm 2 top-hat beam.Further optimisation of the AO mirror profile is expected to lead to an improvement of the FF profile.Figure 5(a) shows the measured temporal pulse profile both at the input and at the output of the main amplifier during 10 J operation.During these experiments no attempt was made to optimize the seed temporal profile to achieve a flat-top output temporal profile.Figure 5(b) shows the output NF measurement at 10 J.As a result of gain saturation in the main amplifier, the output NF cross-section shows a relatively flat-top intensity distribution despite the non-uniformity in the input beam.The lower intensity at the corners of the output NF is due to the very low intensity of the same regions of the input beam profile.Figure 5(c) shows the output FF profile recorded without optimization of the AO mirror.The full-width at half maximum (FWHM) angular spread of the central maximum in the FF image is 137 µrad and 102 µrad along the x-axis and the y-axis, respectively.This corresponds to 3.2 and 2.4 times the diffraction limit for a square 21.5 x 21.5 mm 2 top-hat beam.Further optimization of the AO mirror profile is expected to lead to an improvement of the FF profile.To assess the long-term stability and reliability of the system, the laser was configured to 201 amplify 15 ns pulses to 7 J pulse energy at 100 Hz.Pump and cooling conditions were kept the 202 same as for 10 J amplification, but with a reduced effective seed energy.The laser was operated 203 at this level for 4 hours (corresponding to 1.44 • 10 6 shots) with a long-term rms stability of 1% 204 (Fig. 6a).The data shows periodic output energy oscillations, with a periodicity of about 18 205 minutes, which also characterise the input seed energy (Fig. 6b).This effect is most likely caused 206 by temperature variations in the laboratory affecting the performance of optical components and 207 systems (Fig. 6c).The long-term stability of the laser can be further improved by improving the   To assess the long-term stability and reliability of the system, the laser was configured to amplify 15 ns pulses to 7 J pulse energy at 100 Hz.Pump and cooling conditions were kept the same as for 10 J amplification, but with a reduced effective seed energy.The laser was operated at this level for 4 hours (corresponding to 1.44 • 10 6 shots) with a long-term rms stability of 1% (Fig. 6(a)).The data shows periodic output energy oscillations, with a periodicity of about 18 minutes, which also characterize the input seed energy (Fig. 6(b)).This effect is most likely caused by temperature variations in the laboratory affecting the performance of optical components and systems (Fig. 6(c)).The long-term stability of the laser can be further improved by improving the thermal isolation of PA1 to make it less susceptible to temperature variations and by implementing feedback loop controls to compensate for slow output energy drifts.Throughout the 4 hour run the laser did not show sign of laser-induced damage or component degradation and did not require user intervention.To assess the long-term stability and reliability of the system, the laser was configured to amplify 15 ns pulses to 7 J pulse energy at 100 Hz.Pump and cooling conditions were kept the same as for 10 J amplification, but with a reduced effective seed energy.The laser was operated at this level for 4 hours (corresponding to 1.44 • 10 6 shots) with a long-term rms stability of 1% (Fig. 6a).The data shows periodic output energy oscillations, with a periodicity of about 18 minutes, which also characterise the input seed energy (Fig. 6b).This effect is most likely caused by temperature variations in the laboratory affecting the performance of optical components and systems (Fig. 6c).The long-term stability of the laser can be further improved by improving the thermal isolation of PA1 to make it less susceptible to temperature variations and by implementing feedback loop controls to compensate for slow output energy drifts.Throughout the 4 hour run the laser did not show sign of laser-induced damage or component degradation and did not require user intervention.The polarization uniformity of the output beam was characterized as described in section 2.2.During 7 J, 100 Hz operation, the output beam depolarization is DEP = 6.6%.Figure 7 shows the depolarization patterns at minimum and maximum extinction.No attempt was made to adjust the polarization state of the seed beam at the input of the main amplifier.Optimization of the polarization state of the beam at the input and throughout the multipass [21] and future implementation of depolarization compensation measures [22][23][24][25] is expected to reduce this depolarization level further.
The polarisation uniformity of the output beam was characterised as described in section 2.2.During 7 J, 100 Hz operation, the output beam depolarisation is DEP = 6.6%.Figure 7 shows the depolarisation patterns at minimum and maximum extinction.No attempt was made to adjust the polarisation state of the seed beam at the input of the main amplifier.Optimisation of the polarisation state of the beam at the input and throughout the multipass [21] and future implementation of depolarisation compensation measures [22][23][24][25] is expected to reduce this depolarisation level further.

Conclusion
In this paper we report on the successful demonstration of stable, long-term amplification of 15 ns pulses to 10 J pulse energy at 100 Hz pulse rate (1 kW average power) in a DiPOLE laser.This was achieved with a reduced seed energy of 2.5 mJ by pumping the main amplifier with 39.4 J of total pump energy, corresponding to 25.4% optical-to-optical efficiency.The laser was operated at this level for over 45 minutes (∼ 3 • 10 5 shots) with a 1% rms energy stability.The laser was also operated at 7 J, 100 Hz for 4 hours (corresponding to 1.44 • 10 6 shots) with a long-term energy stability of 1% rms and with 6.6% degree of depolarisation.
Once the room-temperature booster pre-amplifier (PA2) in the front-end section becomes available, it is expected that the increased seed energy will allow full 10 J operation to be achieved with a reduced pump energy, leading to increased efficiency and higher polarisation purity.Also, the quality of the near-field profile is expected to improve as the main cryogenic amplifier will be seeded with a square super-Gaussian profile.Predicted performance of the full system is expected to lead to output energy above 10 J.
These results confirm that DiPOLE technology can be efficiently operated at 100 Hz pulse rate.
Applications of kW-level nanosecond lasers operating at 100 Hz and beyond include laser-shock peening, materials testing and pumping of femtosecond laser amplifiers.
To the best of our knowledge, this is the first time long-term, reliable amplification of a kW-class high energy nanosecond pulsed DPSSL at 100 Hz has been demonstrated.Future work will analyse scalability of the DiPOLE concept beyond the 10 J, 100 Hz level.In previous studies, we scaled the energy from 10 J to 150 J at the pulse rate of 10 Hz [19].We believe that we will be able to apply the lessons learnt to modify the design of DiPOLE-100Hz to achieve higher output pulse energy at 100 Hz.Scalability of pulse rate beyond 100 Hz is subject to ongoing research.

Conclusion
In this paper we report on the successful demonstration of stable, long-term amplification of 15 ns pulses to 10 J pulse energy at 100 Hz pulse rate (1 kW average power) in a DiPOLE laser.This was achieved with a reduced seed energy of 2.5 mJ by pumping the main amplifier with 39.4 J of total pump energy, corresponding to 25.4% optical-to-optical efficiency.The laser was operated at this level for over 45 minutes (∼ 3 • 10 5 shots) with a 1% rms energy stability.The laser was also operated at 7 J, 100 Hz for 4 hours (corresponding to 1.44 • 10 6 shots) with a long-term energy stability of 1% rms and with 6.6% degree of depolarization.
Once the room-temperature booster pre-amplifier (PA2) in the front-end section becomes available, it is expected that the increased seed energy will allow full 10 J operation to be achieved with a reduced pump energy, leading to increased efficiency and higher polarization purity.Also, the quality of the near-field profile is expected to improve as the main cryogenic amplifier will be seeded with a square super-Gaussian profile.Predicted performance of the full system is expected to lead to output energy above 10 J.
These results confirm that DiPOLE technology can be efficiently operated at 100 Hz pulse rate.Applications of kW-level nanosecond lasers operating at 100 Hz and beyond include laser-shock peening, materials testing and pumping of femtosecond laser amplifiers.
To the best of our knowledge, this is the first time long-term, reliable amplification of a kW-class high energy nanosecond pulsed DPSSL at 100 Hz has been demonstrated.Future work will analyse scalability of the DiPOLE concept beyond the 10 J, 100 Hz level.In previous studies, we scaled the energy from 10 J to 150 J at the pulse rate of 10 Hz [19].We believe that we will be able to apply the lessons learnt to modify the design of DiPOLE-100Hz to achieve higher output pulse energy at 100 Hz.Scalability of pulse rate beyond 100 Hz is subject to ongoing research.

Fig. 3 .
Fig. 3. Spatial profile of the seed beam (2.5 mJ at 100 Hz) at the input to the main amplifier.

Fig. 4 .
Fig. 4. Output energy (blue) and optical-to-optical efficiency (red data points) at 100 Hz operation (a).Output energy at 10 J, 100 Hz over ∼ 3 • 10 5 shots, recorded in two separate runs.The vertical dotted line separates the data from the two runs.Inset: shot-to-shot energy stability at 100 Hz (b). 188

Figure
Figure5ashows the measured temporal pulse profile both at the input and at the output of 200

Fig. 3 .
Fig. 3. Spatial profile of the seed beam (2.5 mJ at 100 Hz) at the input to the main amplifier.

Fig. 3 .
Fig. 3. Spatial profile of the seed beam (2.5 mJ at 100 Hz) at the input to the main amplifier.

Fig. 4 .
Fig. 4. Output energy (blue) and optical-to-optical efficiency (red data points) at 100 Hz operation (a).Output energy at 10 J, 100 Hz over ∼ 3 • 10 5 shots, recorded in two separate runs.The vertical dotted line separates the data from the two runs.Inset: shot-to-shot energy stability at 100 Hz (b).

Fig. 4 .
Fig. 4. Output energy (blue) and optical-to-optical efficiency (red data points) at 100 Hz operation (a).Output energy at 10 J, 100 Hz over ∼ 3 • 10 5 shots, recorded in two separate runs.The vertical dotted line separates the data from the two runs.Inset: shot-to-shot energy stability at 100 Hz (b).

Fig. 5 .
Fig. 5. Normalised temporal profiles measured with 4 GHz bandwidth limit (Keysight DSOS404A oscilloscope) of seed pulses (blue) and output pulses (red line) during 10 J, 100 Hz operation (a).Near-field (b) and far-field (c) images of the output beam at 10 J, 100 Hz operation, with cross-section profiles.The colour scale is the same for both images.

208
thermal isolation of PA1 to make it less susceptible to temperature variations and by implementing 209 feedback loop controls to compensate for slow output energy drifts.Throughout the 4 hour 210 run the laser did not show sign of laser-induced damage or component degradation and did not 211 require user intervention.

Fig. 6 .
Fig. 6.Long term stability over 4 hours at 100 Hz, corresponding to 1.44 • 10 6 shots (a).Seed energy at the input of the main cryogenic multipass during the 4 hour run (b).Laboratory air temperature measured over 4 hours, not concurrently with the output and seed energy measurements (c).

Fig. 5 .
Fig. 5. Normalized temporal profiles measured with 4 GHz bandwidth limit (Keysight DSOS404A oscilloscope) of seed pulses (blue) and output pulses (red line) during 10 J, 100 Hz operation (a).Near-field (b) and far-field (c) images of the output beam at 10 J, 100 Hz operation, with cross-section profiles.The color scale is the same for both images.

Fig. 5 .
Fig. 5. Normalised temporal profiles measured with 4 GHz bandwidth limit (Keysight DSOS404A oscilloscope) of seed pulses (blue) and output pulses (red line) during 10 J, 100 Hz operation (a).Near-field (b) and far-field (c) images of the output beam at 10 J, 100 Hz operation, with cross-section profiles.The colour scale is the same for both images.

Fig. 6 .
Fig. 6.Long term stability over 4 hours at 100 Hz, corresponding to 1.44 • 10 6 shots (a).Seed energy at the input of the main cryogenic multipass during the 4 hour run (b).Laboratory air temperature measured over 4 hours, not concurrently with the output and seed energy measurements (c).

Fig. 6 .
Fig. 6.Long term stability over 4 hours at 100 Hz, corresponding to 1.44 • 10 6 shots (a).Seed energy at the input of the main cryogenic multipass during the 4 hour run (b).Laboratory air temperature measured over 4 hours, not concurrently with the output and seed energy measurements (c).

Fig. 7 .
Fig. 7. Depolarisation patterns at maximum (a) and minimum (b) extinction at 7 J, 100 Hz operation.The colour scale is the same for both images.

Fig. 7 .
Fig. 7. Depolarization patterns at maximum (a) and minimum (b) extinction at 7 J, 100 Hz operation.The color scale is the same for both images.