Patent Application
20050271094
The present invention relates to methods and devices for optical parametric chirp pulse amplification method and apparatus for high power optical amplification of ultrashort optical pulses in the infrared wavelength range.
High power ultra-short optical pulses have found numerous applications in the last two decades. Large peak powers of such pulses allowed accessing the highly non-linear regime of light-matter interactions. Laser spectroscopy, material processing, production of deep UV and X ray pulses are several fields that benefited greatly from these developments. The standard technique for production of such pulses is chirp pulse amplification (CPA). A review of this technique and applications of the ultrashort pulses can be found in Perry M D, Mourou G, “Terawatt to Petawatt subpicosecond lasers”, Science, 264 (5161) 917-924 (1994).
After years of development and great success, it is becoming increasingly clear that CPA technique is reaching its limits. Two major ones are gain narrowing and thermal deformations in the laser gain medium. Although there have been some clever improvements have with hollow fiber approaches and cryogenically cooled amplifiers to solve these problems; they don't provide potential for future scaling in power. Another problem with classical OPC systems is that they can provide amplified pulses only within certain wavelengths that in turn depend on the quantum mechanical level structure in the available laser gain materials.
Particularly, a major problem is in generation of ultrashort pulses in the IR spectral region (0.7-20 um). Such pulses have several significant scientific, technological, and medical applications. Many important vibrational transitions in organic molecules (O—H and C—H stretches for example.) or intersubband transitions in semiconductor nanostructures occur in this region. Practical applications of ultrashort pulses in this spectral range occur in medicine such as the ablation of biological tissues or photodynamic therapies. Currently, the prevailing method of generating such pulses involves optical parametric devices pumped by amplified ultrafast Ti:Sapphire laser systems. However, these systems are cumbersome, complicated to operate, and can not provide high power outputs.
In the last several years, an alternative technique for producing high power ultrashort laser pulses has emerged. The principle of optical parametric chirped pulse amplification (OPCPA) was first disclosed in A. Dubietis, G. Jonusauskas, and A. Piskarskas, “Powerful femtosecond pulse generation by chirped and stretched pulse parametric amplification in BBO crystal”, Opt. Commun. 88, 437-440 (1992) which since then has generated a great deal of interest for its potential to produce high energy ultrashort pulses. Dubietis et al disclosed stretching an ultrashort pulse by chirping it (typically ˜100 fs pulse is stretched to 0.1-0.5 ns) then amplifying the pulse in an optical parametric amplifier where it is approximately spatially and temporally overlapped with a high energy pump pulse in the phase matched configuration. After the amplification, the chirped pulse is compressed again to its original duration producing an ultrashort pulse with large energy
General OPCPA design considerations were disclosed later in 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, 125-133 (1997).
Although the method resembles standard chirped pulse amplification (CPA) there are significant differences. The physical interaction in the OPCPA is non-resonant which involves no thermal deposition. This offers an advantage over the conventional CPA technique in which thermal distortions severely limit high average power scaling. Other important advantages include improved pulse contrast, increased amplification bandwidth and higher gain. These advantages were discussed in Ross et al.
To date more then 60 articles have been published in international journals discussing and applying the OPCPA design. Still, the aforementioned OPCPA advantages have not been exploited to date due to several shortcomings. In order for an OPCPA system to become a useful laser tool several conditions must be fulfilled.
The first issue is stability of the output pulse. Interaction of the amplified ultra-short pulse with matter is very non-linear which means small fluctuations of the laser pulse intensity can cause large fluctuations in desired effect. In many applications, intensity stability levels of the amplified pulse have to involve amplitude changes of less than of 1-2%. It is a non trivial problem to produce such pulses from OPCPA's that typically operate in the high gain limit. Small variations of the pump pulse intensity can cause large variations of the output signal intensity. Typically 5-10% variations in stability have been achieved so far. Gain saturation is the standard solution but the saturation point has to be optimized by tuning the input pump and signal intensities. It is not easy to achieve that condition across the whole signal profile in the OPCPA amplifier since typically the temporal and spatial intensity profiles of the input pump and signal pulses are highly modulated.
Secondly, high energy conversion efficiency between pump and signal pulses is very desirable since it contributes to overall system compactness and efficiency. This conversion is largest in the saturation regime that can not be achieved easily. This is also a non trivial issue for the same reason mentioned in the last paragraph.
Thirdly, the amplification in the non-linear medium must preserve enough signal bandwidth to allow compression of the final pulse to short durations. This problem in bandwidth arises in the OPCPA for two reasons. The first one is the phase matching condition that has to be preserved between wave vectors of the three interacting optical waves. This problem is well known. The second reason is the wavelength chirp of the input signal pulse where spectral pulse profile is mapped into the temporal profile of the pulse. Trailing and leading pulse edges receive smaller gain than the central portion of the pulse which results in spectral narrowing of the output pulse.
Finally, the system has to be flexible in terms of choice of signal and pump wavelengths to allow amplification of the desired spectrum to be achieved with the most suitable pump sources.
Optical parametric amplification as a nonlinear process is critically dependent on the input signal and pump intensities. After the signal and pump pulses enter the nonlinear medium there is a periodic exchange of energy between them. At first, the energy transfers from the more energetic pump pulse to the signal. After a certain length the energy starts to flow back from the signal to the pump. This length is called the saturation length. It is dependent on the initial pump and signal intensities and the amount of the pump-signal phase mismatch. Unless the pump and signal optical waves are close to the saturation point the speed of energy transfer is very large. This can cause additional noise in the amplified pulse and reduced conversion efficiency.
In the article by Dubietis et al. discussed above, the original OPCPA technique was proposed. The pump pulse was derived from the signal pulse by using a beam splitter and amplified in a regenerative amplifier. Although this scheme generates well synchronized pump pulses there is no flexibility in choosing the pump wavelength. The same problem exists in the method described in EU Patent CN1297268.
In U.S. Pat. No. 6,181,463 issued to Galvanauskas et al. long pump pulses are used (>1 ns) and triggered electronically to make them in sync with the signal pulse. This approach gives poor conversion efficiency because of the bad overlap between signal and pump pulses. The electronics triggering also introduces timing noise.
In U.S. Pat. No. 6,873,454 issued to Barty et al. a solution for this problem is presented where OPCPA is combined with a classical laser amplifier. Although this approach solves the efficiency and timing problem with electronically triggered long pump pulses it is not flexible in terms of choosing signal wavelength and also is not suitable for amplification of very short pulses because of the problem with gain narrowing in the classical amplifier.
A solution for optimizing intensity pump and signal profiles for OPA interaction is disclosed in Giardalben et al, Optics Express, Vol 11, Issue 20, 2511-2524 (2003), but the solution relies on using fast electro-optics. Such an approach does not give enough control and precision for controlling these profiles. Further, electro-optical components are cumbersome and not used easily.
The present invention provides an optical pulse amplification system,
The present invention also provides a method of laser pulse amplification, comprising the steps of:
The present invention provides a method and apparatus for generating high power ultrashort pulses, preferably in the IR spectral range (0.7-20 Ãm) by using an of optical parametric chirped pulse amplification (OPCPA) system in which pump pulses are produced from a mode-locked laser system synchronized to a mode-locked laser which produces seed laser pulses, both of which are directed to a non-linear material in which energy from the pump pulse is transferred to the seed pulse thereby amplifying it.
The method may use passive or active pre-shaping of the intensity envelopes of the pump pulses before they interact with the signal pulses, or the seed pulses may be modified by active preshaping of the intensity envelope of the seed pulses.
The present invention also provides an optical pulse amplification system, comprising:
These and other objects will be apparent based on the disclosure within.
For a better understanding of the present invention and to show more clearly how it may be carried into effect, reference will now be made, by way of example, to the accompanying drawings which show a preferred embodiment of the present invention and in which:
a) shows the temporal overlap between pump and signal laser pulses in a conventional (prior art) OPCPA method which relies upon a large difference between the pump and signal laser pulse duration to compensate for the lack of timing stability between the two pulses;
b) shows that in the present method, the synchronization of pump and signal mode-locked lasers allows the duration of the two input laser pulses to be within the same order of magnitude, without sacrificing temporal stability of the amplification;
Definitions
A mode-locked laser is a laser that functions by modulating the energy content of each laser resonator's mode internally to give rise selectively to energy bursts of high peak power and short duration in the sub-nanosecond domain.
When we refer at least two mode locked lasers being synchronized this means they are synchronized to each other such that a time delay between arrival of the first stretched seed laser pulse and said pump laser pulse at the nonlinear optical medium fluctuates in time by an amount shorter than pulse durations of the stretched seed laser pulse and said pump laser pulse to give substantially temporally and spatially overlapped stretched seed laser pulse and pump laser pulses in the nonlinear gain media.
By timing jitter we mean random variation in the timing of arrival of laser pulses at a certain point relative to a specified clock. In the present application the clock is defined by a pulse train of a signal mode-locked laser.
Diffractive optics means optical elements that diffract incident laser beam pulses with certain wavelengths under pre-determined specific angles depending on the laser beam wavelength and the point of incidence.
Nonlinear optical material refers to an optical material that possesses a strong nonlinear dielectric response function to optical radiation. The non-linear medium used in the present invention is selected to give energy transfer from the second laser pulses to the first laser pulses through a non-linear optical interaction.
Combining elements are optical elements that direct, and/or shape, and/or focus a laser beam such that it is incident at a determined position with determined size and under determined angle. Common examples include mirrors, lenses, wedges, prisms, wavepleates, polarizers, beam-splitters, filters and any combination thereof, etc. Their usage is well known to people skilled in art.
We disclose two methods for generating well shaped and synchronized short optical pump pulses in the present OPCPA design.
The first method, referred to as synchronization method 1 includes using any mode-locked pump laser system that is actively or passively synchronized to a mode-locked signal laser and produces pulses with durations that are approximately equal to stretched signal pulses before the amplification. Well designed mode-locked laser systems can produce pulses with less then 1% intensity fluctuations.
a) and 1b) show diagrammatically the difference between one of the known prior art OPCPA methods and the method disclosed herein. In these conventional OPCPA methods the pump pulses have nanosecond pulse (>1 ns) durations and are generated from high power laser sources with poor timing control, see
In synchronization method 1 disclosed herein, the pump pulse is generated from a mode locked laser which is passively or actively synchronized with a mode-locked signal laser. There are methods well known to people skilled in art for controlling pulse durations from such lasers by adjusting the mode-locked laser parameters or placing a bandwidth limiting element within the mode-locked laser resonator. Further, there are well-known methods of synchronizing two independent lasers with relative timing jitter much smaller than typical durations of the signal pulses in the OPCPA systems. This allows precise temporal overlap between signal and pump pulses (τs˜τp) in the non-linear medium as shown in
System 10 includes an optical parametric amplifier 22 comprising a nonlinear optical material for receiving the combined pump and signal pulses and amplifying the signal pulses using energy of the pump pulses. The non-linear material possesses a strong nonlinear dielectric response function to optical radiation which gives rise to substantial energy transfer from the pump pulse to the signal pulse through non-linear optical interaction.
The wavelengths λp, λs and λi of the pump, signal and idler beams respectively must satisfy phase matching-conditions:
1/λp=1/λs+1/λi
np/λp=ns/λs+ni/λi
Optimizing the choice and orientation of the non-linear medium to satisfy these conditions is well known to the people skilled in the art.
The nonlinear medium of the parametric amplifier 22 may be any of the following nonlinear crystals; KnBO3, MgO:LiNbO3, BBO, LBO, RTA, KTA, KTP, AgGaSe2, AgGaSe, or any listed in the attached crystal bibliography [6]. The nonlinear medium may be a quasi-phase matched crystal, including periodically poled versions of all crystals listed in [6], e.g., PPLN, PPKTP and PPKTA.
One or both of the pump sources 14 or signal sources 12 may contain a mode locked fibre laser. Specifically where the signal laser can be a high-bandwidth erbium doped fibre laser at 1.5 μm and/or the pump source can contain a Yb-doped fibre laser at 1.0 μm Combining fibre laser technology with parametric amplification to yield compact and robust sources of high-energy IR pulses.
The pump source 14 may include a mode locked solid state rare-earth doped laser or the second or third harmonics of that laser system. The signal source may be a mode locked Titanium Sapphire laser with or without optical absolute carrier phase stabilization, or the second or third harmonics of that laser system.
The output of the non-linear amplifier medium 22 may be useful for some applications by itself. In a preferred embodiment shown in the
Referring to
The combining elements 42 may include one or more diffractive optics as shown in
Extension of this method can include any combination of independent mode-locked laser systems producing pump pulses with different wavelengths which are all synchronized to the same signal mode-locked laser. The output of these pump mode-locked lasers (which could be amplified) can be used for pumping a multistage OPCPA.
Because of phase matching constraints it can happen that the optimal pump wavelength in the OPA is not the same as the one derived from the mode-locked pump laser. In that case pump wavelength can be shifted before OPA by harmonic generation (like SHG, THG etc) in a non-linear medium. This embodiment is shown on
Similarly, it may be beneficial to change the signal pulse wavelength before the pulses enter the OPA. That wavelength can be shifted by using methods well known to the people skilled in art. Examples include SHG, THG, spectrum broadening and/or shifting in fibers etc. The embodiment is represented on the
Systems 60 and 70 can be implemented without optical compressor 24 if a short output pulse duration is not needed.
Synchronization Schemes
Several possible synchronization schemes can be implemented between the signal mode-locked laser 12 and the pump mode-locked laser 14 in the different embodiments shown in
Each of the following synchronization schemes can be used, but the aforementioned OPCPA methods are not limited to them.
Scheme 1
The pump mode-locked laser is actively mode locked by amplitude or frequency modulators where the RF driving signal for these modulators is provided by RF filtering of the fundamental RF frequency or one of its harmonics of the electrical signal from the photo detector observing the pulse train from the signal mode-locked laser. Alternatively, the RF driving field for modulators can be created by electronically dividing or multiplying in integer multiples the fundamental RF frequency or one of its harmonics from the electrical signal output of a photo detector observing the pulse train from said mode-locked laser. In addition, phase locked loops can be employed to reduce the relative timing jitter between said signal and said pump mode-locked lasers. Analog or digital phase detectors detect the phase error between trains of electrical pulses coming from photo detectors observing optical pulse trains from said signal and said pump mode-locked lasers. The phase error is then electronically converted to the phase correction signal applied either on the RF signal coming to the modulators or to the position of the translation stage on which one of the pump laser end mirrors is mounted.
Scheme 2
The variant of the Scheme 1 can be employed where an additional cavity dumping element inside the said pump mode-locked laser is installed. The cavity dumping element dumps the mode-locked pump pulses directly from the resonator which results in larger pump pulse energies. Even larger improvements can be realised by placing additional Q-switching elements inside the said pump mode-locked laser to increase the pump pulse energy
Scheme 3
The pump mode-locked laser is a passively mode locked laser (e.g. by using saturable absorbers or Kerr lens mode-locking). The synchronization with the said signal mode-locked laser is achieved by dynamic control of the pump mode-locked laser cavity length. Analog or digital phase detectors detect the phase error between trains of electrical pulses coming from photo detectors observing optical pulse trains from said signal and said pump mode-locked lasers. The phase error is then electronically converted to the phase correction signal which is then used to control the position of the translation stage on which one of the pump laser end mirrors is mounted. Readjusting the cavity length then takes place until the phase error is minimized.
Scheme 4
The pump mode-locked laser is passively mode locked by using a saturable absorber. The fraction of the mode-locked pulses from the said signal mode-locked laser is converted to pulses with a wavelength that is within absorption spectrum of the saturable absorber. If these converted pulses are made incident on the saturable absorber such that the incidence spot is overlapped spatially with the incidence spot for the intra-cavity pump mode-locked pulse, the pump mode-locked laser dynamics will favour operation when the cavity loss of the said pump mode-locked laser is minimized. This will lead to synchronization of the optical pulse trains from said signal and said pump mode locked oscillators. For converting the energy of the signal mode locked pulses to pulses and wavelengths which are incident on the saturable absorber the techniques mention above can be used.
An alternative method for creating pump pulses well synchronized to the signal pulses can be done by utilizing optical nonlinear processes. This method is referred to as synchronization method 2. Not only does this approach allow to minimize the jitter between pump and signal pulses, it provides a method to control the phase of the idler wave produced through the OPCPA process. It is well known that the phase of the idler wave is sensitive to the phase of the pump and signal pulses. The equations found in reference B. A. Saleh and M. C. Teich, “Fundamentals of Photonics”, Wiley, (1991) chapter 19, pages 762-774 reveal that the rate of growth (along the z-axis) of the idler wave scales proportionally with the complex field amplitude of the pump and signal pulses; those complex field amplitudes include the phase of the pump and signal pulse. This pump pulse could be derived from the continuum generated in a high nonlinearity optical fibre, such as tapered fibres or various forms of microstructure fibres, or fibres made of a highly nonlinear glass material (chalcogenide glasses are one potential example).
The process of continuum light generation in a high nonlinearity fibre provides a pathway to generate a reference, phase-locked optical pulse with respect to the pulse that has generated the continuum. It is generally believed that continuum light is generated as follows in high-nonlinearity fibres, at least for low-intensity femtosecond pulses (this would certainly apply to the case of 100-fs, 1-nJ pulses from a mode-locked erbium-doped fibre laser). The fibre dispersion is such that its second-order dispersion vanishes in the vicinity of the pulse central wavelength. The basic idea is that the fibre dispersion becomes anomalous for wavelengths above 800 nm (up to roughly 1600 nm). The part of an input pulse in the wavelength range with anomalous dispersion becomes a higher-order soliton with number N (N being larger than unity); that N soliton breaks into many fundamental solitons (i.e. of order N=1) This phenomenon was discussed by Hermann et al, Experimental evidence for supercontinuum generation by fission of higher-order solitons in photonic fibres”, Phys. Rev. Lett., vol. 88, paper 173901 (2002).
These fundamental solitons have a spectrum at red-shifted wavelengths with respect to the input pulses, when these input pulses come from a Ti:Sapphire laser emitting around 800 nm; it would be the other way if the pulses come from an erbium-doped fibre laser emitting at 1550 nm. By changing the parameter of the input pulses (duration, energy), one changes the order N of the initial high-order soliton. By changing N, one can change the number of frequency-shifted solitons, and the position of their central wavelength. The generation of frequency-shifted solitons in the anomalous dispersion region is accompanied by the emission of phase-matched nonsolitonic radiation in the wavelength range with normal dispersion. This nonsolitonic radiation takes the shape of optical pulses whose central wavelength is adjusted by the soliton order N. These mechanisms are well-illustrated in
In short, according to this scheme, one could tune a specific optical pulse to the gain center of a regenerative amplifier used to pump the OPCPA. Injection seeding the regenerative amplifier with such a pulse allows for a full control of the timing and of the phase of the pump pulse to be used in the OPCPA.
This description of continuum generation seems to be validated with low-power, long pulses (100-fs duration and above). The key element is to have an input pulse with sufficient power to correspond to a high-order soliton (of order N=2 and above). This can be realized with longer pulses, or with short pulses of higher energy.
If continuum generation (or spectral broadening) is produced according to another mechanism, the same picture of injection seeding the regenerative amplifier with the same source that produces the pulses to be amplified would lock the phase of the idler wave generated through parametric amplification, and would also synchronize the amplifier with respect to the seed source.
It should be noted that the method based on continuum generation (or spectral broadening) in a high-nonlinearity fibre is fundamentally different from the process of soliton self-frequency shift taking place in standard optical fibres currently used for telecommunications. In the process of soliton self-frequency shift the carrier frequency of a single, solitonic pulse shifts to the red due to Raman-type interactions.
The spectrally broadened pulse from 82 is subsequently injected into device 84 where soliton wavelength is selected and duration adjusted to generate desired pump pulse for the amplifier 22. The pump pulse is amplified in the pump amplifier 32 before amplifying the seed pulse in the amplifier 22. The optical compressor 24 can be placed after the amplifier 22 in case if shorter durations of the amplified pulse are desired.
There are many ways of selecting a soliton or a pulse of nonsolitonic radiation produced through continuum generation in microstructured or tapered fibres (the method does not require the seeded pulse to be a soliton; continuum generation also produces pulses of nonsolitonic radiation). Among possible methods, there are listed as flollows. 1) The gain linewidth of the active medium in the pump pulse amplifier. 2) filters introduced in the cavity of the pump pulse amplifiers, such as: a single birefringent filter, or a combination of many birefringent filters; a single prism, or a combination of many prisms; a single thin-film filter, or a combination of many thin film filters; a single Fabry-Perot étalon, or a combination of many Fabry-Perot étalons; other optical interferometers (Michelson, Mach-Zehnder, Fox-Smith, or a combination of many optical interferometers; a single diffraction grating (including holographic gratings), or a combination of many diffraction gratings (including holographic gratings); a single volume hologram, or a combination of many volume holograms; any arrangement combining any of the afore-mentioned filters. 3) Same filters mentioned in 2), but positioned between the nonlinear medium that has produced spectral broadening and the pump pulse amplifier. 4) A single fibre Bragg grating, or a combination of many fiber Bragg gratings, positioned between the nonlinear medium that has produced spectral broadening and the cavity of the pump amplifier.
Optimizing Intensity Profiles of the Pump and Signal Pulses
Optical parametric amplification as a nonlinear process is critically depended on the input signal and pump intensities. After the signal and pump pulses enter the nonlinear medium there is a periodic exchange of energy between them. At first, the energy transfers from the more energetic pump pulse to the signal. After a certain length the energy starts to flow back from the signal to the pump. This length is called the saturation length. It is dependent on the initial pump and signal intensities and the amount of the pump-signal phase mismatch. Unless the pump and signal optical waves are close to the saturation point the speed of energy transfer is very large. These causes critical dependence of output signal level to input pump fluctuations if operating point is away from saturation. This feature is undesirable since it introduces noise.
The other unwanted effect of the high gain in the OPA is bandwidth narrowing. Since the signal pulse is chirped it has temporal modulation that corresponds to its spectrum. The central, more intense part of the signal pulse receives more gain than its wings resulting in the spectral narrowing or the output pulse.
Both problems can be reduced if intensity profiles of the input pump and signal pulses are tuned such that all spatial-temporal points of the signal pulse reach gain saturation at the OPA output approximately simultaneously. Such optimal pump and signal input intensities always exist since three-wave mixing equations that describe the dynamics of the process are deterministic and can be always solved backwards.
Each of the following intensity tuning schemes can be used, but the invention is not limited to them.
Scheme 1
In the first method the signal temporal profile is shaped. Here we use the fact that the signal pulse is chirped and that the signal pulse spectrum is mapped into its temporal profile. In an embodiment of the system shown at 90 in
Scheme 2
In a second method the pump temporal profile is passively shaped. This method for shaping the pump intensity profile is based on passive pre-shaping of the pump pulses in a three wave mixing process in nonlinear medium 102 in
Further, the pump pulse energy shaped intensity profile can be converted to another pump pulse with another wavelength by harmonic generation before it interacts with the signal pulse and said other pump pulse can be recombined with the signal pulse in the parametric amplifier. The intensity profile of the wavelength shifted pulse can be optimized by optimizing the intensity profile of the fundamental pulse. The system for generating pump pulse 92 can be any of aforementioned OPCPA pumping methods but is not limited to them. Also systems 90 and 100 can be used with or without compressor 24 after parametric amplification in the amplifier 22.
As used herein, the terms “comprises”, “comprising”, “including” and “includes” are to be construed as being inclusive and open ended, and not exclusive. Specifically, when used in this specification including claims, the terms “comprises”, “comprising”, “including” and “includes”, and variations thereof mean the specified features, steps or components are included. These terms are not to be interpreted to exclude the presence of other features, steps or components.
The foregoing description of the preferred embodiments of the invention has been presented to illustrate the principles of the invention and not to limit the invention to the particular embodiment illustrated. It is intended that the scope of the invention be defined by all of the embodiments encompassed within the following claims and their equivalents.
List of Acronyms
U.S. Patent Documents
This patent application relates to, and claims the priority benefit from, U.S. Provisional Patent Application Ser. No. 60/570,899 filed on May 14, 2004, which is incorporated herein by reference in its entirety.
| Number | Date | Country | |
|---|---|---|---|
| 60570899 | May 2004 | US |
