MEC provides a range of laser capabilities through two distinct laser systems. Most of the experiments that require high energy pulses in the nanosecond regime are typically performed using the long pulse laser system, whereas research involving high peak intensity lasers, such as high energy density physics, secondary X-ray or particle sources, or sub-picosecond time resolved phenomena use the short pulse laser system. The following diagram represents the laser chain for both of these laser systems.
Figure 1. Schematic diagram of laser system at MEC. (Left) Short pulse laser system. (Right) Long pulse laser system.
Long Pulse Laser System
The long pulse laser at MEC is designed to provide high-energy shapeable pulses at 527 nm. The front end of the system consists of a CW fiber oscillator operating at 1053 nm, which is injected into a programmable electro-optical (EO) modulator and shaped to the desired pulse length and shape. The output of the pulse shaper is fiber-coupled to a multi-pass diode-pumped Nd:Glass amplifier and temporally filtered using a pulse slicer, which minimizes the effects of amplified spontaneous emission in the pre-amplifiers. The pulse is subsequently amplified in ø25 mm Nd:glass rods to the J-level, at a rate of up to one shot per two minutes, and split into four arms. Each arm is further amplified in ø50 mm Nd:glass rods, and finally frequency doubled to 527 nm. In standard configuration, the four arms are polarization multiplexed into two beamlines and expanded to ø 72 mm to be focused to target. Up to 15 J per arm, for total energy of 60 J, can be delivered in a 10 ns pulse.
Figure 1. Schematic of the MEC long pulse laser following a summer 2017 upgrade.
Figure 2. Table layout of MEC long pulse laser.
In the long pulse front end, a 140 mW CW fiber source is modulated and then amplified in diode-pumped Nd:YLF heads, delivering >100 mJ of pulse energy at 10 Hz. The front end pulses are temporally filtered using a pulse slicer to minimize the effects of amplified spontaneous emission in the pre-amplifiers. The pulse is subsequently amplified to the Joule level (up to 1 shot per 2 minutes) and split into four arms. Each arm is further amplified by two 50 mm glass rods (total of eight rods) at a shot rate of 1 per 7 minutes. Finally, the arms are frequency-doubled to 527 nm using large aperture KDP and LBO crystals. Figure 3 shows the results of energy measurements of the two east and two west laser arms indicating the >30 J/arm in the IR with good frequency doubling efficiency to 527 nm. Here, the LBO crystals are shown to double slightly more efficiently than the KDP. After the second harmonic stages, the pulse energy per arm in operation is up to 15 J for 10 ns square pulses, giving 60 J focused to target. The system has demonstrated ablation pressures exceeding 3 Mbar.
Figure 3. Energy scaling of for flat-top temporally shaped pulses. In normal operation, pulse energy is throttled to ~15 J per arm in 10 ns pulses for long term system health.
With a diode-pumped front end replacing a flashlamp pumped system in a 2017 upgrade, the ns laser exhibits excellent stability in energy and spatial profile. The improvement in spatial profile is shown in Figure 4.
Figure 4. The front end spatial modes are shown pre- and post 2017 upgrade. Diode pumping of the front end gives a more stable spatial profile that can be optimized to a smoother beam profile.
In the front end the pulse is modulated using an IX Blue LiNbO3 EO device with a tap for closed-loop bias control, which helps to maintain pulse shape stability. The highland T400 waveform generator driving the modulator is controlled with in-house closed-loop optimization code in Python and updates instantaneously, allowing rapid convergence to a desired pulse shape in the front end. At 5 Hz, convergence is achieved in seconds.
While diverse waveforms are readily generated using the front-end closed loop system, the amplification stages significantly affect the final pulse shape. A good starting point can be established using a semi-empirical transfer function to converge at 5 Hz to the appropriate front end output for a given desired pulse shape, but the finishing touches require a few iterations on full energy shots at 1 shot/7 min.
A variety of waveforms can be generated for experiments. Figures 6 and 7 serve as examples of waveforms utilized in experiments at MEC, measured at full energy at the output of the long pulse laser. Pulse duration can be varied between ~5 and 35 ns. Maximum pulse energy will depend on the chosen shape.
Figure 5. Example delivered 10 ns (top) and 20 ns (bottom) waveforms: flat and slight rise square pulses, compared to target shapes (dashed lines).
Figure 6. Two examples of user-delivered “ramp” pulses compared to target shapes (dashed lines).
Figure 7. Example of user-delivered flat square pulses of different pulse length. These demonstrate roughly the maximum power for each pulse duration when a flat top is required.
Figure 8. Example of user-delivered step pulses with different foot lengths and height, relative to target (dashed lines).
During experiments, the waveform of each arm is monitored separately and recorded for each laser shot, together with its calibrated energy.
Shot to Shot Stability
The upgrade in 2017 to a diode pumped front end has greatly improved the shot-to-shot energy stability of the long pulse laser as compared to the previous system. Example measurements are shown in Figure 9.
Figure 9. Shot to shot variation in an example run, using square pulses. (Left) each variation in energy for each arm. %RMS deviation is (AB, EF, GH, IJ): 1.5, 2.9, 0.75, 1.1; (Right) %RMS variation in total energy over the five shots: 0.9.
Focusing Optics and Far Field Beam Profiles
The two beam lines inside the target chamber are focused with 250 or 500 mm focal length lenses, with the 250 mm aspheric lenses being standard. Typically these lenses are paired with phase plates to produce large supergaussian focal spots of predefined size. For the 250 mm lenses these are nominally 150 mm, 300 mm and 600 mm. Example spot measurements with each of the phase plates are shown in Figure 10, using a log false color scale in the images for contrast.
Figure 10. Focal spot measurements for different phase plates (rows) and beamlines (columns), using the standard 250 mm lenses. The images use a log false color scale for contrast. Radial profiles are shown for each, as well as average intensity and contained energy values for a standard 60 J in 10 ns flat top pulse. No measurements are available using the non-standard 500 mm lenses, but the spatial scales would roughly double, and intensities reduced to 1/4.
Diagnostics for Long Pulse Laser system
The diagnostic toolbox for the long pulse laser on every shot includes characterization of the pulse temporal profile and energy for each arm. The relative timing of the laser beams with respect to the x-ray beams is monitored with a resolution of about 20 ps. The focal spot on target is measured off-line at low energy before the run.
Short Pulse Laser System
The short pulse laser at MEC is a double-CPA laser designed to operate at >25 TW peak power, with < 40 fs pulse duration and > 1 J per pulse. A schematic of the laser system is shown in Figure 1. A commercial CPA (Legend) system, seeded by an oscillator that is RF locked to LCLS, generates mJ-level compressed pulses, which are passed through an OPA pulse cleaning system before being stretched again for further amplification. The first bowtie amplifier operates at 120 Hz, matching the LCLS x-ray pulse rate, outputting >10 mJ per pulse. A second bowtie stage amplifies to >1.5 J at 5 Hz repetition rate. After compression, this yields > 1 J in 40 fs, for 25 TW peak power.
Figure 1. Diagram of the MEC short pulse laser system.
The final output has a pulse duration of approximately 35-55 fs, with a beam diameter of 65 mm, and an energy of 1 J. The compressor grating spacing can be adjusted to produce between 35 fs and 8 ps pulses. For pulses of longer duration, stretched pulses with 150 ps duration have been provided successfully.
Figure 2. (legacy information: please contact) Wizzler measurement of the pulse shape of the MEC short pulse laser indicating compression to nearly transform-limited pulse at 23 nm bandwidth.
After amplification, a 90 mm clear aperture deformable mirror (DM; cf. Figure 3), using feedback from an online wave front sensor (WFS), corrects the distortions produced in the laser chain. Focal spot and wave front measurements in the chamber allow for pre-correction of wave front distortion from the compressor and focusing optics. This system is necessary for producing the highest intensity at focus.
Figure 3. A photo of the deformable mirror of the MEC short pulse laser is shown.
|Before Correction||After Correction|
|RMS||0.6 µm||22 nm|
|P-V||2.5 µm||100 nm|
Table 3. Demonstrated wavefront correction of the DM/WFS closed loop system.
Short pulse experiments use non-standard beam delivery, so users should contact MEC scientists to discuss the best focusing options and expected performance for their experiment. An inventory of off axis parabolic mirrors (OAPs) is given in table 4. Note that the standard full-power (1 J, <50 fs) beam size is roughly 65 cm, and so OAP 4 is incompatible with this mode. A recent example measurement of the focus from OAP 3 is given in Figure 4.
|Name||CA (mm)||Effective focal length(mm)||Off axis angle (deg)||coating||notes|
|OAP1||99||311.66||17.55||Dielectric||Incompatible with standard target mount|
|OAP2||99||311.66||17.55||Dielectric||Incompatible with standard target mount|
|OAP3||99||330.4||35.31||Silver||Primary OAP for full intensity|
|OAP4||50||500.58||34.41||Silver||For long focal length and low power. Only compatible with reduced beam size and power.|
Table 4.The OIPs owned by MEC. OAP 3 is used in the standard short pulse delivery platform. The standard full energy beam size is 50 mm.
Nearly all user experiments using the multi-TW beam require excellent contrast to be successful. To achieve highest baseline pulse contrast, the short pulse laser system employs a double-CPA front end. After initial chirped pulse amplification of the oscillator pulses at 120 Hz, the beam is recompressed and passed through a nonlinear filter (with an OPA and SFG-based pulse cleaner  replacing XPW starting in Run 17) and then stretched again for final amplification. This filter removes prepulses and pedestal energy that builds up primarily in the laser front end. Measurements utilizing third order autocorrelation techniques revealed a contrast better than 107 in the > 3 picosecond range using a double CPA pulse cleaning system. To further increase this contrast, MEC can provide a 2nd harmonic stage that, in combination with dichroics, should theoretically achieve better than 1014 contrast beyond 3 ps, while sacrificing pulse energy (output ~300 mJ at 400 nm).
Timing Jitter and Synchronization with the LCLS X-ray Beam
Taking full advantage of the temporal resolution of femtosecond X-ray pulses and femtosecond optical lasers in pump-probe experiments requires timing measurement of a few 10 fs or better. LCLS has a pulse-to-pulse timing jitter relative to the accelerator radio-frequency (RF) distribution of approximately 60 fs RMS, integrated over a bandwidth of 0.1-100 kHz. The MEC short pulse laser oscillator is locked to the accelerator RF, with a distribution with similar or better timing jitter. Drifts in the laser beam path and RF distribution need to be controlled to approximately the 1 ps level.
The mode-locked seed laser oscillator operates at 68 MHz, the seventh sub-harmonic of the 476 MHz RF reference frequency of LCLS, which defines the fine (sub-picosecond) timing of laser pulses in the absence of configuration changes to the laser system.
The noise performance of the LCLS laser locking system was measured to have an RMS laser-to-RF-reference timing jitter of 25 fs between 100 Hz and 100 kHz. Below 100 Hz, phase noise is dominated by the noise of the linac RF reference.
To surpass the limitations of this jitter, experiments involving the fastest physical phenomena (e.g. transitions of core shell electrons) measure the relative timing between the laser and X-rays on a shot-by-shot basis by means of X-ray/optical cross-correlation techniques, referred to as the time tool, pictured in Figure 5. Ultrafast ionization by the x-ray beam in YAG windows allows its synchronization relative to the MEC laser to be precisely set. The ionization causes the window to become opaque showing definite before/after timing. By angling a window relative to the x-ray beam, the time tool gives synchronization to within a few fs. The time tool is cross-referenced to the target plane using another YAG window at the laser focus position, with a small f-stop temporarily inserted to spread the laser focus at the target plane. The time tool measures an RMS jitter of ~ 100 fs between the MEC laser and the x-ray beam.
Figure 5. Image of the MEC timing tool.
Figure 6. Synchronization of laser to the X-ray the time tool. Top: laser shadography of a target window on two consecutive laser shots, arriving just before (left) and just after (right) an X-ray pulse (square shaped beam); bottom: The time tool shows arrival time of the ionizing X-ray pulse inline, demonstrating that the synchronization jitter between the two consecutive pulses shown here is less than 100 fs.
Diagnostics for the Short Pulse Laser system
The short pulse laser diagnostic suite includes a variety of devices that are accessible on-line and on demand. Here is a list of the common diagnostics already installed in the toolbox.
- On-shot diagnostics:
- Time tool (synchronization)
- Pulse energy
- Near-field beam profile
- Equivalent plane focal spot
- On-demand diagnostics (off-line)
- 3rd order autocorrelation
- Far-field beam profile
- Motorized power meters
 “Pulse contrast enhancement via non-collinear sum-frequency generation with the signal and idler of an optical parametric amplifier,” vol. 114, no. 22, p. 221106, Jun. 2019.