The technique of optical parametric chirped pulse amplification (OPCPA) [1
I. N. Ross, P. Matousek, M. Towrie, A. J. Langley, and J. L. Collier, “The prospects for ultrashort pulse duration and ultrahigh intensity using optical parametric chirped pulse amplifiers,” Opt. Commun.
144(1-3), 125–133 (1997). [CrossRef]
A. Dubietis, G. Jonusauskas, and A. Piskarskas, “Powerful femtosecond pulse generation by chirped and stretched pulse parametric amplification in BBO crystal,” Opt. Commun.
88(4-6), 437–440 (1992). [CrossRef]
] enables the amplification of broad bandwidth laser pulses. This technique opens the possibility of very high peak-power laser systems with promising practical designs of laser architectures with peak powers exceeding 1 PW. This paper discusses a novel front-end system based on OPCPA and designed to seed a 10PW laser system that will produce pulses with 300 J of energy in 30 fs and capable of generating focused intensities greater than 1023
. The development of this front-end has been conducted to demonstrate that a suitable seed can be generated at the required wavelength and to study the characteristics of such a system. To that end the system reported here has been constructed using equipment readily available to prove the principles that will be employed.
To achieve the combination of 300J in 30fs large beam diameters are required at the final stage of pulse amplification to prevent damage to optical components. Consequently gain media with diameters > 150 mm are required. The technique of OPCPA is an attractive approach with non-linear crystals of KDP and KD*P being commercially available with suitable dimensions. Of the two it has been shown that KD*P has the larger gain bandwidth in high energy applications. Figure 1
shows the calculated gain bandwidth for a small signal gain of 100 in KD*P pumped at 527nm in a non-collinear geometry. As can be seen from Fig. 1
OPA gain over large spectral range of ~160 nm can be achieved at a center wavelength of 910 nm [3
V. V. Lozhkarev, G. I. Freidman, V. N. Ginzburg, E. V. Katin, E. A. Khazanov, A. V. Kirsanov, G. A. Luchinin, A. N. Mal'shakov, M. A. Martyanov, O. V. Palashov, A. K. Poteomkin, A. M. Sergeev, A. A. Shaykin, and I. V. Yakovlev, “Compact 0.56 petawatt laser system based on optical parametric chirped pulse amplification in KD*P crystals,” Laser Phys. Lett.
4(6), 421–427 (2007). [CrossRef]
V. V. Lozhkarev, G. I. Freidman, V. N. Ginzburg, E. A. Khazanov, O. V. Palashov, A. M. Sergeev, and I. V. Yakovlev, “Study of broadband optical parametric chirped pulse amplification in DKDP crystal pumped by the second harmonic of a Nd:YLF laser,” Laser Phys.
15, 1319–1333 (2005).
Fig. 1 Gain bandwidth for a non-collinear interaction in KD*P pumped at 527nm.
To seed a laser system based on KD*P requires a broad bandwidth pulse centered at 910 nm. This lies at a portion of the Ti: Sapphire gain curve which would be unsuitable and other gain materials have insufficient bandwidth in this region. At focused intensities of 1023
the contrast of the laser pulse becomes crucial for the successful application of the laser. With plasma formation occurring on solid targets at 1012
contrast ratios of 1011
are required for the peak of the pulse to the nanosecond background fluorescence introduced during amplification. To that end we have developed a novel large bandwidth front-end system suitable for seeding a high peak-power OPCPA system. A schematic overview of the design of the front-end system is shown in Fig. 2
. The initial task is to generate a pulse with sufficient bandwidth to support a 30fs pulse these pulses are then amplified to 100s of μJs on the picosecond timescale. This limits the temporal window of the parasitic fluorescence that seeds the rest of the system to the duration of the pump pulse ~10 ps. This clean pulse is then stretched to 1.87ns before further amplification to the joule level. In this way the multi-stage OPCPA front-end is able to satisfy such criteria and is ready to be used for further development of the 10 PW system. In this paper we will discuss the different components in turn and their influence on the output of the system.
Fig. 2 Schematic of the front-end.
2. Seed generation and mJ OPA stages
The initial seed is generated using picosecond timescale OPCPA to limit the duration of the initial parametric fluorescence that seeds the rest of the system. To generate the broad bandwidth required the technique of chirp compensation is employed [5
K. Osvay and I. N. Ross, “Broadband sum-frequency generation by chirp-assisted group-velocity matching,” J. Opt. Soc. Am. B
13(7), 1431–1438 (1996). [CrossRef]
]. This scheme operates in a collinear non-degenerate geometry. This geometry is instantaneously narrowband. However, for the same phase matching angle there are a number of combinations of signal, pump and idler wavelengths that satisfy the phase matching criteria. This process was described in detail in our recent publication [6
Y. Tang, I. N. Ross, C. Hernandez-Gomez, G. H. C. New, I. Musgrave, O. V. Chekhlov, P. Matousek, and J. L. Collier, “Optical parametric chirped-pulse amplification source suitable for seeding high-energy systems,” Opt. Lett.
33(20), 2386–2388 (2008). [CrossRef] [PubMed]
] and forms the first stage of the system which combined with an additional 2 stages increase the seed energy to 500μJ. The scheme requires the relative chirps between the signal and pump to be well controlled. In our case we have a phase matching angle of 31°, a signal chirp of 17.5 nm/ps and a pump chirp of 1 nm/ps. This enables us to generate pulses that have a bandwidth of 140 nm.
A schematic of the three-stage system is shown in Fig. 3
, the seed and pump pulses are formed by spectrally dividing the output of a Ti:sapphire laser using a spectral beam splitter. A portion of the spectrum with 150 nm bandwidth centered at 720 nm is used as the seed whilst a 20 nm section of the spectrum at 800 nm is used to form the pump beam. The pump beam is stretched in a grating stretcher to ~10 ps before being amplified to 1.8mJ in a multipass amplifier (Femtolasers Compact Pro). This amplifier operates at a 1KHz repetition rate which is synchronized to the oscillator pulses using an RF signal derived from a photo-diode. The amplified pump pulse is then frequency doubled to 400 nm in BBO1. The seed pulse is stretched in a prism stretcher to ~10 ps. The pump and seed beams are then mixed in crystal LBO1 amplifying the seed and generating an idler signal centered at 910 nm and having ~140 nm FWHM spectrum as shown in Fig. 4
. There is a small non collinear angle between the pump and signal beams to enable the separation of the signal and idler beams. The pump fluence for this stage is 150mJ/cm2
and the incident seed energy is 0.5nJ. The energy contained in the idler after this first stage is ~2 µJ and forms the seed beam for the subsequent stages.
Fig. 3 Broadband pulse generation and mJ level OPA stages layout.
Fig. 4 Spectrum of the idler generated at the first stage of mJ level OPA.
The undepleted pump from the first stage is relayed to LBO2 to act as the pump for the 2nd stage. The 910nm idler from the 1st stage is relayed to form the seed of the 2nd stage with the 720nm beam from the 1st stage blocked in the near-field between the 2 stages. The pump fluence in the second stage is 170mJ/cm2 and the output pulses have 25 µJ retaining the bandwidth.
To further increase the energy a 3rd stage is used. The pump beam is formed by amplifying the unconverted 800nm from BBO1 in an additional Ti:sapphire amplifier. This amplifier is a 3-pass bow-tie geometry pumped from both ends and increases the pump pulse energy for stage 3 to ~30 mJ at 800 nm. A relatively highly doped Ti:sapphire crystal with an absorption coefficient of ~5.5 cm−1 at 532 nm and a single pass absorption of ~89% was chosen to be used as the gain medium of the amplifier. The length of the Ti:sapphire crystal is 4 mm and the diameter of the crystal is 7 mm. The crystal has uncoated Brewster cut faces at both ends, so that for 10ps pulses it can withstand over 30 mJ of infrared (IR) at the maximum fluence of ~1 J/cm2. A commercial pulsed Nd:YAG laser (CFR400) with internal second harmonic conversion is used as the pump laser for this amplifier. The CFR400 is capable of delivering laser pulse energies up to ~230 mJ at 532 nm at a repetition rate of 10 Hz. The 10Hz triggering signal is derived by dividing the KHz amplifier triggering signal by 100. The output pulse energy of ~30 mJ in the IR was obtained at a pumping level of ~205 mJ with an overall gain factor of ~35. It was observed that both the fluorescence and amplified IR beam profile are quite uniform across the gain medium which is a critical and important feature of the pump beam for efficient OPA amplification. The spectral bandwidth of the IR seed pulse was maintained due to negligible gain narrowing in this amplifier. The amplified signal is then frequency doubled to 6.5 mJ at 400 nm pulse for pumping the third OPA stage. This final stage generates 500 µJ pulses with the 140nm bandwidth maintained.
To demonstrate that the output of this system can be compressed a simple single pass compressor was established for the 910nm output. This comprised of a single block of 40.5mm SF10 and a grating compressor with 1200l/mm gratings mounted at Littrow angle for 910nm with a separation of 9.5mm and an out-of-plane angle of 12°. A single shot autocorrelator was used to measure the pulse to be 25fs. The contrast of the seed was then measured using a Sequoia. The Sequoia is a commercial high dynamic range third-order cross-correlator produced by Amplitude Technologies [7
]. The contrast trace is shown in Fig. 5
. As can be seen the dynamic range achieved with this combination of energy and pulse duration was 5 orders of magnitude and the contrast of the pulse is 105
at −1.2ps. The shoulder that the pulse sits in that goes from −1ps to 0.5ps is attributed to incomplete compression because the shape of this region could be influenced by the separation of the grating compressor. Despite this shoulder the contrast measured at this point is better than that observed at the output of the system as discussed later.
Fig. 5 Contrast of the compressed output from the seed generation scheme.
3. Seed pulse stretching
To evaluate the performance of an all OPCPA front-end system amplifying a broad band seed pulse to the Joule level a test CPA scheme was designed to increase the pulse length to a comparable length to that required to be compatible with a kJ level pump laser. The length of this stretch for a single CPA scheme will be governed by the optimal pulse duration for extraction of energy from the final pump laser, typically several nanoseconds. The pulse stretching is obtained by using a double pass stretcher [8
I. N. Ross, A. J. Langley, and P. Taday, “A simple achromatic pulse stretcher,” in Central Laser Facility Rutherford Appleton Laboratory Annual Report 1999–2000 (Council for the Central Laboratory of the Research Councils, 2000), pp. 201–203.
] which has been modified from our previous experiments [9
O. V. Chekhlov, J. L. Collier, I. N. Ross, P. K. Bates, M. Notley, C. Hernandez-Gomez, W. Shaikh, C. N. Danson, D. Neely, P. Matousek, S. Hancock, and L. Cardoso, “35 J broadband femtosecond optical parametric chirped pulse amplification system,” Opt. Lett.
31(24), 3665–3667 (2006). [CrossRef] [PubMed]
] to increase its bandwidth to 140nm limited by the size of the available gratings. A schematic of the stretcher and the seed pulse injection is presented in Fig. 6
Fig. 6 Schematic of the stretcher injection.
The pulse is incident at an angle on grating 1 which has a grove density of 1480 lines per mm and is tilted to be at the Littrow angle for 900nm. The pulse is then directed to a focusing mirror f = 500 mm where the beam is off axis. The center of the radius of curvature of this mirror is such that it is vertically at the center of the gratings but placed horizontally between them. After the mirror the pulse is incident on gating 2 before being retro-reflected by the plane mirror which is placed close to the image plane of the mirror. To minimize the astigmatic aberrations introduced by the stretcher it requires a beam waist between the first grating and the curved mirror. The control of the pulse injection to the stretcher and the output pulse rejection is achieved using a polarizing beamsplitter cube and the combination of a Faraday rotator and a half wave plate. The stretcher demonstrates ~50% transmission efficiency delivering ~1.87 ns output pulses. There is a spectral loss of the output pulse spectrum due to size of the second grating in the stretcher introducing hard clips at 830 and 970nm.
4. Joule level OPA stages
After stretching the pulse is imaged and expanded into the first crystal of the Joule level OPA amplification scheme. The amplification for this section is achieved in two stages as schematically presented in Fig. 7
Fig. 7 Joule level OPA stages layout.
The setup involves usage of 19 mm and 20 mm long LBO crystals respectively for each stage. The crystals have been placed into electrically heated metal ovens mounted on 4-axis stages. The OPA process is non-collinear type-1 geometry, where the seed and idler have the same polarization (set as P) and the pump is orthogonal to this (set as S).
The seed is imaged from LBO1 to LBO2 using a vacuum relay telescope which also magnifies the beam from 2 to 10mm in diameter. An optical delay line is used in the seed path to provide timing adjustment between the interacting pulses in LBO2.
The pump beam line has been built up with three vacuum telescopes to provide collimated and image relayed beams of about 4 mm in diameter on the LBO1 and 10 mm on LBO2. The pump laser delivers 4.5 J pulses at 532 nm at a 2 Hz repetition rate with a square temporal profile and spatial top hat profile (Continuum custom laser). The 2 Hz trigger signal is generated by dividing the 1KHz trigger signal by 500 and re-synchronizing this signal to the oscillator RF. A combination of 90/10 beam splitter, half wave plates, polarizers and beam dumps has been arranged to provide pump energy control for both stages of amplification and deliver up to ~3 J/cm2 pump fluence on each crystal. The undepleted pump beams are blocked after each crystal.
At a pump fluence of 3J/cm2
the maximum small signal gain (SSG) of stage 1 was measured to be ~400 and produced over 12 mJ of output seed energy for 20μJ of input seed energy. Further amplification in stage 2 generated 0.5J of output energy at a pump fluence of 2.5J/cm2
with a SSG of 40. Evolution of the amplified pulse spectrum in the Joule level OPA stages is shown in Fig. 8
. As mentioned previously the initial seed spectrum (blue trace) is limited by the size of the gratings in the stretcher. Further loss in spectrum of the amplified pulse is caused by the bandwidth of the broadband mirrors, although providing ~100 nm of bandwidth is achieved at the output.
Fig. 8 J level OPCPA output pulse spectra.
5. Pulse compression
To test how compressible the amplified pulse is, the output beam from the Joule level OPA stages was injected into a 4-grating double-pass compressor. A schematic of the plan view of the compressor is shown in Fig. 9
. This geometry was imposed on the compressor due to the size of the available gratings and limits the bandwidth of the compressed pulse to 60nm.
Fig. 9 Schematic of the plan view of the compressor.
The four gold gratings have a line density of 1480 l/mm and are used at Littrow angle in an out-of-plane design. It is double passed using a roof mirror to displace the beams horizontally on the return path. Typical near and far-field images are shown in Fig. 10
Fig. 10 Typical near (a) and far (b) field images.
The pulse duration of the compressed pulses was measured using a SPIDER to be 35 fs. The SPIDER operates in the far-field so there is no information as to any spatial variation in the pulse duration. Figure 11(a)
shows the temporal shape of the pulse and Fig. 11(b)
shows the measure spectral intensity and phase of the compressed pulse.
Fig. 11 Pulse duration of the final compressed pulse measured by the SPIDER technique. The temporal shape is shown in (a) and the spectral intensity (solid) and phase (dashed) are shown in (b).
The contrast of the compressed output has been measured using a combination of photo-diodes and a Sequoia. The diode system had a dynamic range of 10 orders of magnitude and showed that there was no parametric fluorescence within this range on the nanosecond time scale. The contrast trace measured using the Sequoia is shown in Fig. 12
. As can be seen there are a number of features, the pulse sits on a shoulder that starts at 30ps before the main pulse. This is due to fluorescence from the 3rd stage of the seed generation scheme; further work is required to determine why it starts at 30ps. The coherent contrast appears to start at −10ps This feature could be due to scattering in the stretcher reported elsewhere [10
C. Hooker, Y. Tang, O. Chekhlov, J. Collier, E. Divall, K. Ertel, S. Hawkes, B. Parry, and P. P. Rajeev, “Improving coherent contrast of petawatt laser pulses,” Opt. Express
19(3), 2193–2203 (2011). [CrossRef] [PubMed]
Fig. 12 Contrast measurement of the close in contrast.
In conclusion we have constructed an all OPCPA front-end scheme for seeding large aperture high peak-power OPCPA systems. We have demonstrated the generation of a seed source at 910nm with at least 140nm of bandwidth. These seed pulses have been generated and undergone initial amplification on the picosecond timescale to the sub milli-joule level and compressed to 25fs. An evaluation CPA system has then enabled these pulses to be further amplified to the Joule level and then compressed to 35fs. The aperture of the gratings used in the stretcher and compressor prevented the preservation of the entire bandwidth generated in the milli-joule system.
The contrast of these pulses has been measured at two points in the system. The measurement of the background fluorescence indicates that it is better than 10 orders of magnitude, limited by the dynamic range of the photodiodes. We are confident that it is sufficient to meet the required ratio of 1011 with respect to the peak of the compressed pulse. Comparison of the measurements of the coherent contrast after the milli-joule stage and the joule stage shows that the contrast of the initial seed is better than that after final compression. Further work will be undertaken to confirm if this is due to scattering in the stretcher by investigating alternative stretcher geometries and if other modifications to the stretcher can be made to improve the contrast.