Double plasma mirror for ultrahigh temporal contrast ultraintense laser pulses

September 18, 2017 | Autor: Anna Levy | Categoria: Optics, Quantum Physics, Optical physics, High Power, Electrical And Electronic Engineering, Peak Power
Share Embed


Descrição do Produto

310

OPTICS LETTERS / Vol. 32, No. 3 / February 1, 2007

Double plasma mirror for ultrahigh temporal contrast ultraintense laser pulses Anna Lévy, Tiberio Ceccotti, Pascal D’Oliveira, Fabrice Réau, Michel Perdrix, Fabien Quéré, Pascal Monot, Michel Bougeard, Hervé Lagadec, and Philippe Martin Service des Photons, Atomes et Molécules, Commissariat à l’Energie Atomique, Centre d’Etudes de Saclay, 91191 Gif sur Yvette, France

Jean-Paul Geindre and Patrick Audebert Laboratoire pour l’Utilisation des Lasers Intenses, UMR7605, CNRS, Commissariat a l’Energie Atomique, Ecole Polytechnique-Paris 6, 91128 Palaiseau, France Received September 1, 2006; revised October 27, 2006; accepted November 2, 2006; posted November 3, 2006 (Doc. ID 74695); published January 12, 2007 We present and characterize a very efficient optical device that employs the plasma mirror technique to increase the contrast of high-power laser systems. Contrast improvements higher than 104 with 50% transmission are shown to be routinely achieved on a typical 10 TW laser system when the pulse is reflected on two consecutive plasma mirrors. Used at the end of the laser system, this double plasma mirror preserves the spatial profile of the initial beam, is unaffected by shot-to-shot fluctuations, and is suitable for most high peak power laser systems. We use the generation of high-order harmonics as an effective test for the contrast improvement produced by the double plasma mirrors. © 2007 Optical Society of America OCIS codes: 140.7090, 190.4160, 320.7080, 350.5400.

With the introduction of the chirped pulse amplification (CPA) technique,1 very high intensity laser pulses, with subpicosecond duration and tens of terawatt peak power, are now available on relatively small size tabletop laser systems. Properly focusing such an intense pulse allows one to get intensities of up to 1022 W / cm2.2 However, a drawback of this technique is the presence of a temporal pedestal on the nanosecond time scale [mainly due to amplified spontaneous emission (ASE)] and prepulses on a subpicosecond time scale (due, for instance, to imperfect matching of some optical elements inside the laser chain). The magnitude of these pulse components is usually expressed through the contrast value, which is defined as the ratio between the laser peak and background intensities. As the peak intensity routinely exceeds 1018 W / cm2, the pedestal is sufficient to create a preplasma that expands in front of the original surface of the target. As a consequence, the main pulse will not interact with a steep density gradient, which prevents, for instance, the efficient generation of well-collimated high-order harmonics (HOH) from solid surfaces.3,4 Among the different methods developed to remove the pedestal, a particularly efficient one is based on the self-induced plasma shuttering, or plasma mirror (PM), technique.5,6 Using this technique at the Saclay Laser Interaction Center (SLIC, France), we have demonstrated an improvement of the temporal contrast by a factor of 100 of a 10 TW Ti:sapphire laser system (600 mJ, 60 fs CPA system), leading to an overall contrast higher than 108 and an energy transmission of 70%.7 This “clean” laser pulse has been used to generate HOH by nonlinear reflection on a plasma with a steep electronic density gradient.8 Well-collimated harmonics up to 20th order have been observed for a laser intensity of 3.1017 W / cm2, while no harmonics were ob0146-9592/07/030310-3/$15.00

tained without using the plasma mirror technique. This technique then allows the study of the interaction of high-intensity ultrashort pulses with solid targets in situations that are free of plasma expansion. Maintaining the same setup, we switched to a tighter focus using a f / 2.5 parabola 共f = 300 mm兲 and a PM position adjusted to keep the same fluence, intended to yield an intensity of a few 1018 W / cm2. Under those conditions, no HOH could be detected. In fact, the generation of HOH strongly depends on the state of the target surface. The fluence in the pedestal of the pulse must then stay below the damage threshold of the target, i.e., 40 J / cm2 in the nanosecond regime,9 for the dielectric targets used in these experiments. As the ASE pedestal lasts about 2 ns and the main pulse about 65 fs, in the setup without PM ( 7 ⫻ 105 contrast), the energy contained in the pedestal is about 5%. With the f / 6 off-axis parabola 共f = 500 mm兲, the focal spot measured 30 ␮m radius leading to a total fluence of 20 kJ/ cm2. Therefore the pedestal value is well above the damage threshold of the sample, preventing any efficient HOH generation. A single PM (SPM) increases the contrast ratio by 2 orders of magnitude. In those conditions, the pedestal fluence does not exceed 10 J / cm2, below the damage threshold, and HOH generation becomes possible. With an f-number of f / 2.5, the pedestal fluence on the target reached 80 kJ/ cm2. Thus the contrast obtained with a SPM is not sufficient for experiments in this range of intensities (1018 W / cm2 or higher). This example shows how increasing the peak intensity requires a parallel increase of the temporal contrast. Typically, 1020 W / cm2 onto the target will require a contrast at least equal to 1010. In this Letter, we describe the implementation and the characterization of a double plasma mirror (DPM). Using two consecutive reflections on a set of antireflection© 2007 Optical Society of America

February 1, 2007 / Vol. 32, No. 3 / OPTICS LETTERS

Fig. 1. (Color online) Experimental setup for the DPM (the laser beam comes from the left).

coated (AR) parallel dielectric plates,8 we expect a gain of 104 and a transmission of 50% in the optimum configuration, when considering the performance of a single PM. These two PMs can be used in several ways. One possibility is to insert them between the target and the short focal length optics used to focus the beam on the target. Such a setup is compact but is not very flexible and can be challenging to implement. In addition, both PMs work in the intermediate field, leading to hardly predictable amplitude distortion at the focal point. We have chosen to implement the PMs before the optics that focus the beam on the target, inserting them in a confocal setup. The constraints on the phase and amplitude distortions induced by the second PM are less severe, since the beam reflected from this mirror will be imaged on the final target. The position of both PMs with respect to the focal spot determines the quality of the DPM, in terms of contrast gain and energy transmission, as well as the quality of the focal spot. Once the laser pulse has been compressed to 65 fs, the laser beam is sent in the vacuum chamber housing the DPM (see Fig. 1). The DPM vacuum chamber is ideally placed between the compressor and the experimental chamber. It may be interesting to compare the results of a given high-field experiment with and without the effect of the pedestal. To that aim, a pair of retractable mirrors placed inside the DPM chamber allows, or not, the beam to pass through the DPM. The laser beam is first focused by a dielectric f = 1230 mm off-axis parabola between the two plasma mirrors. The position of the focal spot between the two PMs can be varied using an external motorized stage. After reflection on each PM, the beam is collimated by a second identical off-axis parabola and sent as a parallel beam in the experimental chamber. The setup is maintained under clean vacuum down to at least 10−6 mbars. The laser beam is s-polarized with respect to each PM surface to maximize the PM reflectivity7 and the distance between them is about 4 cm. We use an antireflection coating deposited on a dielectric bulk material (BK7; 150 mm⫻ 40 mm⫻ 20 mm) and whose spectrum ranges from 750 to 850 nm. To facilitate the beam alignment procedure, a small metallic coating area is left on one end of both mirrors. After each shot, the DPM is automatically shifted to offer a fresh surface for the next shot. This system allows about 1800 shots before the replacement of both mirrors is required, and it can operate around 1 Hz. In the case of subpicosecond pulses, the efficiency of a DPM is affected by the incident fluence on each PM7 in only a minor way. Typically, PMs can accommodate

311

fluences between 50 J / cm2 and 1 kJ/ cm2 without a significant distortion of the wavefront for reflectivities ranging from 70% to 85%. Numerical simulations suggest that the second mirror could be triggered at higher fluences (which results in a slightly higher reflectivity) because the contrast is higher. However, the interplay among the initial wavefront structure, the wavefront distortions in the intermediate field, and the plasma temperature can make predictions difficult. So we measured the energy transmission of the DPM for three different positions of the focal spot between the two PMs: halfway between the PMs and about 1 cm away from each PM. Then, for the best focal position, we have determined the temporal contrast and the focal spot size. To achieve these tasks, the laser beam is focused, after the DPM, in the center of the interaction chamber using an f / 3.6 off-axis parabola 共f = 300 mm兲. We used a joule meter to get the transmission with and without the DPM. To measure the focal spot, we placed a 16-bit CCD camera just after an imaging relay system. We measured the temporal contrast with a third-order cross correlator (SEQUOIA, Amplitude Technologies, France) placed just after a recovery system. The high dynamics of this device 共1012兲 is obtained by varying the detector (photomultiplier) high-voltage power supply and by using a set of calibrated neutral-density filters. Its temporal resolution is 120 fs. We had previously found a transmission of 50% when the focus is placed halfway between the two PMs and 40% for the other two positions. The best result is then obtained for the halfway focal position, which results in a fluence of 50 J / cm2. Note that the conditions are similar to the ones used for the SPM discussed in Ref. 8. The associated contrast is shown in Fig. 2, where each point is an average over 3 shots. For comparison, the results obtained with a SPM mirror8 and without any correction are also presented. The contrast of the UHI10 laser is initially around 105 at 2 ps and 7 ⫻ 105 at long delays. With the DPM, the signal amplitude for long delays (ASE) starts from 1010 and grows by 8 orders of magnitude in 2 ps (coherent contrast region). With the SPM, for the same delays we find 108 and 106. It is remarkable

Fig. 2. (Color online) Temporal profiles of the laser in the DPM configuration (circles), using a SPM (triangles), and without correction (curve).

312

OPTICS LETTERS / Vol. 32, No. 3 / February 1, 2007

that for the transmission, as for the contrast, the DPM behaves exactly as a “SPM at the power 2” (70%2 = 50%, 1002 = 104). We believe that even if this performance could be slightly improved (by varying the distance between the two plates, for instance), this configuration is quite close to the optimum. Note that the temporal profile also shows a modification of the postpulse side. This behavior, as predicted by the modelization,7 is attributed to the plasma expansion, which after a few picoseconds produces a significant defocalization of the postpulse. The image of the focal spot is reported in Fig. 3. The width is identical with and without the DPM. We conclude that the DPM does not introduce any significant distortion of the beam wavefront. In our focusing conditions, the FWHM is approximately 8 ␮m. The peak intensity of the pulse measures 6.1018 W / cm2 for a contrast higher than 1010. As discussed earlier, because the damage thresholds of metals are lower than those of dielectrics, a possible extremely severe test of the laser performance described above consists of performing an HOH experiment from a solid metallic target (gold). Then, if the fluence remaining in the pedestal is larger than a few J / cm2, harmonics cannot be observed. In principle, in our conditions the fluence in the pedestal should be close to 1.2 J / cm2 共6 ⫻ 1018 W / cm2 ⫻ 10−10 ⫻ 2 ns兲. The laser is p-polarized on the gold target, and the reflected beam is sent in an extreme UV spectrometer. A typical spectrum is shown in Fig. 4. A very high level of well-collimated UV harmonic signal is observed up to the 23rd order. This harmonic emission has its origin in the so-called coherent-wake-emission mechanism.10 We noticed an excellent shot-to-shot reproducibility of the spectra. We believe that the absence of preplasma, causing hardly controllable instabilities in the high intensity laser propagation, can partly explain this reproducibility. Using a DPM, we are now able to get routine contrasts higher than 1010. This makes it possible to

Fig. 3. DPM.

(Color online) Spatial intensity profile after the

Fig. 4. Harmonics spectra at 6 ⫻ 1018 W / cm2 and 1010 contrast.

expose, for instance, dielectric solid targets to intensities as high as 1020 W / cm2, without any preplasma induced by the pedestal and prepulses. We believe that with this technique a number of phenomena, such as laser acceleration of particles from solid targets or production of very energetic attosecond pulses, will be accessible to many laboratories. Financial support from the Conseil Général de l’Essonne (ASTRE) program is acknowledged. T. Ceccotti’s e-mail address is [email protected]. References 1. D. Strickland and G. Mourou, Opt. Commun. 56, 219 (1985). 2. S. W. Bahk, P. Rousseau, T. A. Planchon, V. Chvykov, G. Kalintchenko, A. Maksimchuk, G. A. Mourou, and V. Yanovsky, Opt. Lett. 29, 2837 (2004). 3. A. Tarasevitch, A. Orisch, D. von der Linde, P. Balcou, G. Rey, J.-P. Chambaret, U. Teubner, D. Klopfel, and W. Theobald, Phys. Rev. A 62, 023816 (2000). 4. U. Teubner, G. Pretzler, T. Schlegel, K. Eidmann, F. Förster, and K. Witte, Phys. Rev. A 67, 013816 (2003). 5. H. Kapteyn, M. Murnane, A. Skoze, and R. Falcone, Opt. Lett. 16, 490 (1991). 6. D. Gold, Opt. Lett. 19, 2006 (1994). 7. G. Doumy, F. Quéré, O. Gobert, M. Perdrix, P. Martin, P. Audebert, J. C. Gauthier, J.-P. Geindre, and T. Wittmann, Phys. Rev. E 69, 026402 (2004). 8. P. Monot, G. Doumy, S. Dobosz, M. Perdrix, P. D’Oliveira, F. Quéré, F. Réau, P. Martin, P. Audebert, J.-C. Gauthier, and J. P. Geindre, Opt. Lett. 29, 893 (2004). 9. B. C. Stuart, M. D. Feit, A. M. Rubenchik, B. W. Shore, and M. D. Perry, Phys. Rev. Lett. 74, 2248 (1995). 10. F. Quéré, C. Thaury, P. Monot, S. Dobosz, P. Martin, J.-P. Geindre, and P. Audebert, Phys. Rev. Lett. 96, 125004 (2006).

Lihat lebih banyak...

Comentários

Copyright © 2017 DADOSPDF Inc.