METHOD AND SYSTEM FOR HIGH POWER PARAMETRIC AMPLIFICATION OF ULTRA-BROADBAND FEW-CYCLE LASER PULSES

Method and system for high power parametric amplification of ultra-broadband few-cycle laser pulses

FIELD OF THE INVENTION

The present invention relates to parametric amplification. More specifically, the present invention is concerned with a method and system for high power parametric amplification of ultra-broadband few-cycle laser pulses.

BACKGROUND OF THE INVENTION

Generally speaking, the major issues inherent to current amplification methods and system are the opposing requirements for ultra-broadband versus high power amplification. Parametric amplification is currently realized either in a broadband, low gain arrangement referred to as type I amplification, or in a narrowband, high gain configuration referred to as type II amplification [Cerullo03]. Due to phase matching restrictions, a compromise between high amplification factors and reduced amplified bandwidth has to be accepted.

High power amplification of ultra-broadband laser pulses is currently achieved using Optical Parametric Chirped Pulse Amplification (OPCPA) [Dubietis06] using Type I crystals or periodically poled crystals to meet conditions for quasi phase matching, respectively. On the other hand, Type I amplification in a non-collinear geometry (NOPA) can be achieved over broader bandwidths than type II at the expense of lower pulse energies. In both cases the maximum amplification bandwidth is limited. Although OPCPA is a promising approach, capable to support 2 cycle pulses with 740 µJ, this method requires development of new pump sources and its temporal pulse contrast is impaired by significant superfluorescence back ground [Gu09].

In current parametric amplification methods, amplification is carried out in time domain where all spectral components are present at once or in time sequentially in case of OPCPA. The document JP H05 257176 relates to a system and a method for generating an ultra high bit rate light signal train efficiently at a high speed and to perform even the parametric amplification of the signal light at the time of spatial filtering execution on a Fourier transformation surface.

There is still a need in the art for a method and a system for high power amplification of ultra-broadband few-cycle laser pulses in the framework of parametric amplification.

SUMMARY OF THE INVENTION

More specifically, in accordance with the present invention, there is provide a system for high power parametric amplification, comprising at least a first dispersive-and-collimator set; an optical amplification medium located in the Fourier plane of the first dispersive-and-collimator set; and a second dispersive-and-collimator set symmetric to the first dispersive-and-collimator set on an opposite side of said Fourier plane or a reflector at said Fourier plane; wherein a first optical Fourier transformation of a seed beam from a seed source is achieved at the first dispersive-and-collimator set, parametric amplification by overlap of the seed beam with a pump beam from a pump beam source in a spatially dispersed frequency plane using the optical amplification medium occurs in the Fourier plane; and a second optical Fourier transformation is achieved at the second dispersive-and-collimator set, or at the first dispersive-and-collimator set after back refection at the reflector.

There is further provided a method for high power parametric amplification, comprising a first optical Fourier transformation of a seed spectrum and a parametric amplification by overlap of the seed beam with a pump beam in a spatially dispersed frequency plane using an optical amplification medium in the Fourier plane.

Other objects, advantages and features of the present invention will become more apparent upon reading of the following non-restrictive description of specific embodiments thereof, given by way of example only with reference to the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

In the appended drawings:

• Figure 1 shows a schematical view of a system according to an embodiment of an aspect of the present invention;
• Figure 2a shows a schematical top view of the system of Figure 1; Figure 2b shows experimental spectra for ultra-broadband seed and type II amplification;
• Figure 3 shows a side view of a four-pass system according to an embodiment of an aspect of the present invention;
• Figures 4 show a) and b) a temporal characterization of a 2.5 cycle input seed pulse; c) and d) an amplified pulse; and
• Figure 5 is a flowchart of a method according to an embodiment of an aspect of the present invention.

DESCRIPTION OF EMBODIMENTS OF THE INVENTION

Figure 1 illustrates a system according to an embodiment of an aspect of the present invention.

A seed beam spectrum, such as ultra-broadband, carrier envelope phase (CEP) stable seed pulses 10 available in different ways as described below, is passed through a dispersive-and-collimator set, comprising for example a first grating 30 and a first cylindrical mirror 60 as a collimator, for a first Fourier transform into the spectral domain, and overlapped together with a pump beam 40 in the Fourier plane (F), where an optical amplification medium, comprising for example several parametric crystals 50 placed side by side, is provided. In the case of crystals as amplification units of the amplification array, each crystal can be tuned individually for optimum phase matching of its corresponding portion of the broadband input seed bandwidth.

The spectrum is then returned, by a reflector (HR seed and Pump) for example in the case of a folded system, to the first dispersive-and-collimator set, i.e. the grating 30 and mirror 60, for a second Fourier transform back to the time domain back.

In the case of an unfolded system in absence of a reflector, the spectrum is returned, after amplification in the Fourier plane (F), to a second dispersive-and-collimator set (not shown) symmetric to the first dispersive-and-collimator set (30, 60) on the opposite side of the Fourier plane (F).

In Figure 1, the seed beam spectrum is first sent to a pulse shaper 20, such as a Dazzler, prior to the dispersive-and-collimator set, for fine tuning of spectral phase and amplitude to optimize pulse duration and shape after amplification. Furthermore, the pulse shaper 20 has recently been demonstrated to enable CEP controls of titanium doped sapphire amplifier [Canova09].

Collimation of colors can be alternatively achieved by replacing the cylindrical mirror 60 by a second grating with same groove density than the first grating 30 to provide parallel travel of colors on the amplification plane. The spread of colors could be achieved with transmission gratings or prisms as well.

The dispersive element 30 can be a reflection grating, a transmission grating, a prism or a grism.

It has been found that separate nonlinear crystals 50 in the Fourier plane (F) can be optimized to maximize amplification for their corresponding wavelength regime. For instance, several efficient narrowband type II BBO (Beta-Barium Borate) crystals 50 can be utilized to amplify an entire octave of frequencies.

A spatially matching pump beam 40 can be applied in a noncollinear way to allow straight forward beam separation or by use of dichroic optics to combine the beams collinearly. The pump beam may also be provided collinearly using dichroic cylindrical mirrors 60.

A detail of a Fourier plane configuration is given in Figure 2a, in the case of a folded system. Optical imaging by the cylindrical mirror 60 takes place only parallel to the dispersion plane, which is perpendicular to the grooves of grating 30. By this, each color is focused to a line focus whose height is equal the crystal height (perpendicular to the drawing in Figure 2a) and is defined by the input beam diameter.

For low energy seed pulses, ranging between about 1 nJ and 100 µJ, the cylindrical mirror 60 can be replaced by a spherical mirror which focuses each color into a point focus to achieve higher intensity.

As shown in Figure 2a, the spectral components of the collimated seed beam are horizontally dispersed at the grating 30 and travel parallel after the cylindrical mirror 60. Each component is focused to a line in the Fourier-plane (F) and returned to the grating 30 and mirror 60 by the reflector (HR seed and Pump).

The reflector for back reflection of the seed and pump can be a one piece reflector, as shown for example in Figure 1. Alternatively, the reflector can be made of a plurality of separate reflectors, such as individual mirrors 82 as shown in Figure 2a.

Thus, in the Fourier-plane (F), several narrowband nonlinear crystals 50 can be used to amplify corresponding spectral slices. Spectral phase shaping can be carried out by such mirrors 82, which can be translatable (see arrows (A) in Figure 2a). In case the reflector is a one piece reflector and thus not allowing spectral phase shaping as individual reflectors 82 would do, adjustable compensation plates 80 may be used to ensure temporal and spatial overlap of all components. The compensation glass plates 80 can be tilted (see arrow (B) in Figure 2a) to correct for the different thicknesses of the crystals 50 and thereby adjusting the phase delay between the different colors. The compensation glass plates 80 also allow spectral phase shaping of the output pulse. Note that in Figure 2a both mirrors 82 and compensation glass plates 80 are shown, but that compensation glass plates 80 are not necessary in cases when the reflector (HR seed and Pump) is made of individual mirrors 82. Note that a reflector made of a deformable mirror, instead of several mirrors 82, may be used to optimize different phase delays in different regions of the spread spectrum in the Fourier plane.

In any case, by tuning each amplification units 50, such as crystals, phase matching angle by tilting the crystal 50 (see arrow (C) in Figure 2a), not only amplification gain but also amplitude shaping can be realized.

To increase amplification, both pump 40 and seed beam 10 may be made to interact a second time in the optical amplification medium 50, by being back reflected (see high reflector HR Seed and Pump in Figures 1 and 2a in case of a one-piece reflector; see mirrors 82 in Figure 2a in case of a multi-pieces reflector), to achieve two-stage amplification without additional components.

The present system can also be operated in a single pass configuration without back reflection in the Fourier plane. In this case, a second dispersive-and-collimator set (not shown) is placed symmetric to the first dispersive-and-collimator set (30, 60) as shown on the left hand side of the Fourier plane in Figure 2a to the right hand side of the Fourier plane to perform the second optical Fourier transformation.

Experimental spectra for ultra-broadband seed and type II amplification via the phase matching angle of crystals are shown in Figure 2b to illustrate octave spanning type II amplification with five crystals.

Alternatively to type II crystals, type I phase matching crystals can be utilized as well as periodically poled ones in the Fourier plane to enhance the spectral width per millimeter.

Instead of different crystals in the Fourier plane (F), the optical amplification medium array may comprise one crystal with spatially varying phase matching properties, like for instance a spatially varying periodicity in case of periodically poled crystals.

Instead of parametric crystals, real level pumped laser materials, such as real level pumped laser crystals, such as for instance titanium doped sapphire (Ti:Al₂O₃) or neodymium-doped yttrium aluminum garnet (Nd:Y₃Al₅O₁₂), can be used in the same manner.

Alternatively. to solid state amplification media, liquid or gaseous amplification media could be utilized as well

The optical amplification medium may also comprise fiber lasers.

In general, spatial dimensioning of the pump beam can be achieved with an anamorphotic telescope to attain spatial overlap with the spread spectrum in the Fourier plane (see step 60 Figure 5). Maximal temporal overlap is achieved by setting the pump beam duration equal to the one of the seed pulse which is expected to lie in the picosecond regime for the dispersed beam (see step 60 Figure 5). Its duration is given by the vertical focal width in the Fourier plane which depends on focal length of the cylindrical mirror 60 as well as the groove density of the grating.

To combine pump and seed beam, the cylindrical mirror 60 in Figure 2a could be a dichroic mirror which transmits pump and reflects seed beam or an additional beam splitter could be inserted after the cylindrical mirror 60 for combination (step 70 Figure 5).

Alternatively, the pump beam can be supplied under a small angle with respect to the seed as shown in more detail in Figure 3.

The beam path in Figure 2a denotes a double pass through the system, which can be extended to a four pass as shown in Figure 3.

In Figure 3, the weak input seed beam 10 enters from the left handside. Reflection grating 30 and cylindrical mirror 60 are shown as transmission optics for clarity. After this dispersive-and-collimator set (30,60), the seed beam passes the optical amplification medium 50 twice in a first round trip and is back reflected by a first high reflector (see HR pump and seed) towards the grating 30 under a small angle. Those time domain pulses are then send back by a second high reflector (HR seed) into the system.

Correspondingly, the pump beam 40 which passes the cylindrical mirror 60 is back reflected by a third high reflector (see HR pump) to overlay with the twofold amplified beam in time and space. Thus, by adding one reflector for the seed beam (HR seed) and one reflector for the pump beam (HR Pump), four amplification stages can be realized with one set of nonlinear crystals 50.

As mentioned above a symmetric system, i.e. with back reflection at the Fourier plane (F) or with a second collimation and dispersive set on the opposite side of the Fourier plane (F), adds no additional dispersion to the seed beam.

However, the system can also be used as stretcher or compressor, respectively, depending on how the grating separation differs from the focal length [Martinez87]. Thus, large amounts of dispersion that exceed the capability of the pulse shaper 20 (see Figure 1) can be precompensated, if necessary.

A further point to be addressed is the question how to generate a single-cycle optical pulse. One recently demonstrated option is described in [Krauss09], where single-cycle IR pulses have been generated by synchronizing two different fiber oscillators yielding nJ level pulses. Another promising approach is based on the sub two-cycle pulse compression scheme by nonlinear propagation in a hollow-core fiber (HCF) (see [Schmidt10] for example). The experimental input seed spectrum 10 shown in Figure 2b supports a single cycle pulse.

The experimental input seed spectrum 10 shown in Figure 2b supports a single cycle pulse. With the HCF approach, the ability to achieve spectral bandwidths that support single-cycle pulses has been already demonstrated in the laboratory, as shown by the black curve in Figure 2b. Experimentally achieved bandwidths are presented in Figure 2b where the black curve corresponds to a broadened supercontinuum in a HCF [Schmidt10] which could serve as seed spectrum. The narrower BBO2, BBO1 and BBO3 amplification spectra correspond to type II amplification (amplification factor = 18 for a single pass) with a crystal thickness of 3 mm. This schematic plot illustrates that octave spanning amplification is feasible with only five relatively thick type II BBO's.

An operational first prototype setup has been built on the 100Hz laser beam line at the Advanced Laser Light Source (ALLS) of INRS, Varennes, Québec, Canada. The ultra-broadband seed beam centered around 1.8 µm wavelength was provided by spectral broadening of narrowband pulses from a commercial optical parametric amplifier (OPA) in a HCF. After this fiber, 190 µJ pulses were sent to the system which had a total transmission of about 55% yielding 100 µJ at the output without parametric amplification in the Fourier plane. For temporally characterizing the system, a second harmonic generation - frequency resolved optical gating (SHG-FROG) trace was acquired (see Figure 4a). It showed the possibility to transmit a pulse of 16 femtosecond duration at full width at half maximum (FWHM) denoting 2.6 optical cycles at a central wavelength of 1.8 µm. Pumping was carried out with temporally stretched 800 nm pulses whose spatial beam distribution was adapted to ensure perfect beam overlap of seed and pump beams in the Fourier plane. The pump pulses were stretched to few picosecond duration to achieve optimal temporal overlap and to enable higher pump energy compared to pumping with short, femtosecond pulses.

The result of full amplification, i.e. with all crystals tuned for their corresponding wavelength range, yielded 1.5 mJ pulse energy and was characterized temporally and spectrally showing the same bandwidth and duration as the input beam. Remarkably, the FWHM duration did not increase and was measured to be 15 fs, which is equivalent to 2.5 cycles, as shown in Figure 4c. Carrying energy of 1.5 mJ, this system demonstrated a 15 fold increase of energy. This gain factor denoted an efficiency of 11% for converting pump photons to infrared (IR) photons. These numbers can be compared with other OPCPA in the IR range developed in the recent years at MIT in USA (4.5 optical cycles, 0.8 mJ, conversion efficiency 6.5%) and Max Planck Institute of Quantum Optics in Germany (2.2 cycles, 0.74 mJ).

Figure 5 is a flowchart of a method according to an aspect of the present invention.

The spectral phases of a seed beam may be optimized, as discussed hereinabove (steps 10-30), before the spectrum is dispersed (step 40) and collimated (step 50). Then it is overlapped with the pump beam in the Fourier plane (step 70), where crystals of an amplification array are tuned individually for optimum phase matching of their corresponding portion of the broadband input seed bandwidth (step 80). The phase delay between the different colors, due to the different thicknesses of the crystals for example, can be compensated for (step 90). To increase amplification, both pump and seed beam may be made to go through the crystal twice (step 100). The beam is then transformed back to the time domain by a second optical Fourier transform (step 110).

Collimation (step 50) may be done using a cylindrical mirror, a second grating with the same groove density than the first grating, a dichroic mirror, transmission gratings, prisms, a cylindrical mirror followed by a beam splitter, or, for low power pulses, a spherical mirror, for example.

In the time domain, different options for nonlinear spectroscopy are possible. Single- and few- cycle pulses may be generated (step 120).

Moreover, additionally to simple pulse shapes, arbitrary pulse shapes may be generated (step 130) and ultimately, applied for pump-probe or coherent control experiments.

Also, instead of utilizing the entire input spectrum, two different parts of the amplified spectrum may be selected and recombined at the output to perform nonlinear wave mixing, such as for instance difference frequency generation (DFG), which offers the possibility of passive CEP stabilization of the DFG pulse (step 140).

As people in the art will now be in a position to appreciate, the present system and method is based on performing amplification in a frequency domain after time domain pulses are optically Fourier transformed, to overcome bandwidth limitations. In a nutshell, a first optical Fourier transformation of a seed spectrum is performed and parametric amplification is carried out in this spatially dispersed frequency plane. As a consequence, individual parts of the spectrum can be amplified using an optical amplification medium comprising a series of optical amplification units, such as different narrowband crystals, placed one next to each other. Each crystal is tuned independently to optimize its corresponding spectral slice. A second optical Fourier transformation recovers the time domain laser pulses. This method enables scalability of amplified bandwidth and pulse energy at the same time.

The present method and system allow single- and few- cycle pulse amplification without superfluorescence background up to the multi mJ level and beyond. The amplification limit is given only by the damage threshold of the gratings since there the beam profile is smallest. Recently, a damage threshold of 1J/cm² was reported for 40fs pulses [ICFO] which implies the scalability up to the Joule level without additional beam expanders prior to the grating.

On the one hand, the present method and system allow increasing the amplification bandwidth compared to type I amplification and, on the other hand, they allow simultaneous up scaling of pulse energy by increasing the number of amplification crystals but still using small and thick, i.e. of large gain per pass, crystals, thus leading to an ultra-broadband, high power amplification.

The present system and methods can be applied not only in the IR spectral range but also in other wavelength ranges, only limited by transparency of nonlinear crystals and availability of pump lasers. Especially, single-cycle pulse amplification in the visible is feasible when pumping with the second harmonic of femtosecond Chirped Pulse Amplification (CPA) lasers at 800 nm wavelength.

Additionally, the present system allows performing difference frequency generation (DFG) at the output by recombining only two sides of the spectrum. By doing so, a CEP unstable input seed pulse can be passively CEP stabilized via intrapulse DFG.

Amplification in the Fourier domain offers several advantages compared to amplification in the time domain as known in the art.

First, as mentioned above, the general problem of gain narrowing is overcome by distributing a large spectral bandwidth over many amplification crystals. Thus, a number of highly efficient narrowband crystals can be used to amplify unprecedented bandwidths.

Second, simultaneous up scaling of bandwidth and energy can be achieved by providing a larger extension of the Fourier plane. For a given spectral bandwidth, and fixed, still small aperture of crystals, i.e. of about 2mm or less for example, the throughput energy can be increased simply by introducing stronger dispersion of the grating 30 combined with a larger mirror 60. When the Fourier plane (F) extends further, more small crystals are needed. This denotes a huge advantage compared to manufacturing large aperture crystals.

Third, a very good input spatial mode, resulting from propagation in a HCF, is preserved during amplification even with poor pump beam quality. This is because spatial impurities in the frequency plane transfer to homogeneous intensity variations of the spatial mode in the time domain upon optical Fourier transformation. Thus, excellent focusability of the spatial mode is expected without subsequent spatial filtering.

Fourth, the present system intrinsically offers the possibility to perform amplitude and phase shaping throughout the entire bandwidth, i.e. over about one optical octave. Amplitude shaping is achieved by tuning the crystal phase matching angle. Phase shaping in the reflection mode of Figure 2a for example can be carried out by translating the individual high reflectors 82 (HR) in the Fourier plane (F), or by using a deformable mirror. The resolution depends on the number of mirrors, which can be higher than the number of crystals to increase phase shifting accuracy. In combination to mirrors or in case of a high energy transmission setup, respectively, phase shifting can be achieved by rotation of additional glass plates 80.

Fifth, any superfluorescence generated in the Fourier plane is discriminated from the amplified beam due to its different propagation properties, since superfluorescence copropagates with the pump beam, and is not recombined with the seed beam by the grating. Thus, this amplification method provides pulses free of superfluorescence background.

Sixth, arbitrary pulse shapes generated by the pulse shaper 20 (Figure 1) can be amplified without damaging the crystal and without much nonlinear reshaping upon amplification.

The present invention provides a number of advantages, including, for example, scalability of pulse power and spectral width at the same time; utilization of highly efficient narrowband type II nonlinear crystals for amplification; absence of superfluorescence in the output beam; pumped by established CPA techniques; possibility of extracting the conjugated beam given proper design of a second cylindrical mirror and grating; a system inherently enabling additional pulse shaping opportunities.

Although the present invention has been described hereinabove by way of embodiments thereof, it may be modified, without departing from the nature and teachings of the subject invention as recited herein.

References

Cerullo03

G. Cerullo and S. De Silvestri, "Ultrafast optical parametric amplifiers," Rev. Sci. Instrum., 74, 1, 2003.

Dubietis06

A. Dubietis, R. Butkus, and A. P. Piskarskas, "Trends in Chirped Pulse Optical Parametric Amplification", IEEE J. Sel. Top. Quantum Electron., 12, 163 (2006).

Weiner95

A. M. Weiner, IEEE Prog. Quantum Electron. 19, 161 (1995).

Vozzi07

C. Vozzi, F. Calegari, E. Benedetti, S. Gasilov, G. Sansone, G. Cerullo, M. Nisoli, S. De Silvestri, and S. Stagira, "Millijoule-level phase-stabilized few-optical-cycle infrared parametric source", Opt. Lett. 32, 2957 (2007).

Gu09

X. Gu, G. Marcus, Y. Deng, T. Metzger, C. Teisset, N. Ishii, T. Fuji, A. Baltuška, R. Butkus, V. Pervak, H. Ishizuki, T. Taira, T. Kobayashi, R. Kienberger, and F. Krausz, "Generation of carrier envelope-phase-stable 2-cycle 740-µJ pulses at 2.1-µm carrier wavelength," Opt. Express 17, 62-69 (2009).

Canova09

L. Canova, X. Chen, A. Trisorio, A. Jullien, A. Assion, G. Tempea, N. Forget, T. Oksenhendler, and R. Lopez-Martens, "Carrier-envelope phase stabilization and control using a transmission grating compressor and an AOPDF," Opt. Lett. 34,1333, (2009).

Schmidt10

B. E. Schmidt, P. Béjot, M. Giguère, A. D. Shiner, C. Trallero-Herrero, E. Bisson, J. Kasparian, J. -P. Wolf, D. M. Villeneuve, J. -C. Kieffer, P. B. Corkum, and F. Légaré, "Compression of 1.8 µm laser pulses to sub two optical cycles with bulk material", App. Phys. Lett. 96, 121109 (2010).

Krauss09

G. Krauss, S. Lohss, T. Hanke, A. Sell, S. Eggert, R. Huber and A. Leitenstorfer, "Synthesis of a single cycle of light with compact erbium-doped fibre technology" Nature Photonics, 4, 33 (2009).

Martinez87

O. E. Martinez, "3000 times grating compressor with positive group velocity dispersion: application to fiber compensation in 1.3-1.6 µm region", IEEE Journal of quantum electronics, Vol. QE-23, No. 1, 1987.

ICFO

http://www.ibsen.dk/highpower/ibsen/attachments/icfo-test-report.pdf/download.