RU2657305C2 - Method for forming multi-pulse packets of femtosecond laser pulses - Google Patents

Abstract

FIELD: laser engineering.

SUBSTANCE: method of forming laser pulse packets consists in repeated separation of a laser pulse into two pulses that are delayed in time with respect to each other and then combined back. Separation and integration of pulses is performed by means of fiber dividers, and the delay between pulses happens due to their propagation between dividers along fiber channels of different lengths. Use of several pulse separation stages and their relative delay and a combination of such stages allows receiving packets with a different number of laser pulses in them.

EFFECT: technical result consists in ensuring the amplification of a packet of chirped pulses with enabling their identical polarization and increasing the minimum allowable distance between pulses.

4 cl, 6 dwg

Description

FIELD OF THE INVENTION

This invention relates to laser technology and can be used for processing materials by laser radiation.

State of the art

Currently, pico- and femtosecond lasers are increasingly used in laser micro-processing. As a rule, pulses with energies from several microjoules to hundreds of microjoules are necessary for the tasks of microprocessing materials. Moreover, for some materials, it becomes necessary to obtain packages (groups) of ultrashort laser pulses that are separated from each other in time. Typically, the repetition rate of packets (groups) of pulses should be from several kilohertz to several megahertz. The number of pulses in a packet usually does not exceed ten, since a further increase in the number of pulses (while maintaining the total laser energy per pulse packet), as a rule, does not lead to a significant increase in the efficiency of the processing process, and therefore it is not practical. The distance between adjacent pulses should be at least a dozen picoseconds and not more than a few nanoseconds. Amplification of ultrashort laser pulses to the required energies, as a rule, is carried out according to the amplification of chirped pulses using a regenerative amplifier. The problem of obtaining packets of amplified ultrashort laser pulses can be solved by at least two different approaches. One scheme involves amplification of a single ultrashort laser pulse, with its subsequent “multiplication” by several pulses forming a packet. Another approach is to form a packet of low-energy ultrashort pulses at the output of the laser generator (before amplification) and to amplify the entire pulse packet as a whole.

So, in US patent 2006/0018349 A1 of 2006.01.26, a method for generating sequences of laser pulses is disclosed, which makes it possible to obtain sequences of amplified chirped laser pulses at the output of a regenerative amplifier amplifying single laser pulses. The method uses the principle of amplification of chirped pulses with amplification in a regenerative amplifier with a linear resonator, which allows it to be used when working with pulses of picosecond and femtosecond duration. To obtain a packet of laser pulses at the output of the regenerative amplifier, stage-by-stage unloading of the resonator is used due to a quasi-step change in voltage at the Pockels cell. The stepwise voltage path is realized due to the electrical switching circuit. In this method, the possibility of forming various envelopes of the packet of output pulses is substantially limited by the fundamental monotonicity of the control voltage path on the Pockels cell, which does not allow, for example, to “skip” individual bypasses of the resonator and form “complex” envelopes. The distance between adjacent pulses in the packet is determined by the length of the resonator and for typical resonators in such schemes is up to several tens of nanoseconds. The minimum distance between adjacent pulses in a packet is limited by the length of the resonator (since it contains optical elements for controlling the radiation and amplifying it) and the speed of the high-voltage pulse generators available at the current level of technology that control Pockels intracavity cells; this value is several nanoseconds.

US patent 2007/0047601 A1 of 2007.03.01 describes a method for propagating laser pulses using a looped loop optical delay line in a block design containing Pockels cells and amplifying the active element. The delay line loop is connected to the input optical signal through the input and output dividers. The output divider is oriented in such a way that it provides partial transmission of the s-polarized wave. The first Pockels cell injects an input pulse into the delay line. As a result of each round trip along the delay line, a part of the laser pulse leaves it through the output divider and thus forms a separate laser pulse in the output packet. The amplifying element compensates for losses in the delay line so that the output pulses have the same energy. After a number of rounds of the feedback line, the laser pulse circulating along it is emitted using the second Pockels cell. The length of the loop of such a delay line, namely, it determines the distance between pulses in the output packet, must be large enough so that the necessary elements can be placed inside. In addition, this method itself is quite complicated, since it requires the use of both a laser active element and electro-optical crystals, the operation of which is impossible without the corresponding subsystems - pumping and high-voltage sources.

In US patent 5293389 from 1994.03.08 another method is considered - the formation of a group of laser pulses from a single laser pulse. The patent describes a method and apparatus for generating a sequence of time-equidistant laser pulses through the use of a looped passive optical delay line with the conversion of the polarization state. To obtain a packet of pulses, a single laser pulse is supplied to the polarization divider. One part (transmitted or reflected) is sent to the further part of the circuit (working channel) and becomes the first pulse in the generated packet, and the other (on the contrary, reflected or transmitted) falls into the optical delay line. The delay line is looped, i.e. its output is fed to the same polarization divider. This line contains an element that changes the state of radiation polarization (depending on the particular implementation, into a circular, elliptical, other linear or non-polarized state). Due to the change in polarization, the laser pulse passing through the ring of the delay line is re-divided by the polarizer - one part enters the working channel and becomes the second pulse in the packet, and the other returns back to the delay line. This part propagates along the delay line and again partially enters the working channel in the form of a third pulse in the packet. As a delay line, a segment of a fiber waveguide with input and output collimators, or a block mirror optical scheme can be used. In the second case, a phase plate is used as an element controlling radiation depolarization. To adjust the relative energy of the first pulse in the packet between the laser source and the "breeding" device can also be installed phase plate.

In the patent CN 103560391 dated 2014.02.05, another method for generating sequences of laser pulses is proposed in which the propagation of laser pulses occurs due to the difference in the speeds of the pulses with different polarizations when propagating in a crystalline birefringent material. The separation of the initial pulse into pulses with different polarizations is achieved due to the mutual orientation of the optical axes of the birefringent crystalline elements and polarizers located one after another. The described method has a fundamental drawback - as a result of the propagation of radiation through birefringent crystals, the pulses can be delayed relative to each other for a period of not more than 10 ps. A further increase in the delays between pulses can be realized through the use of crystalline elements of a very large length or in a very large quantity, however, this significantly increases the cost of implementing such a solution. To further increase the distance between pulses (up to nanoseconds), the authors of 103560391 use additional lines and delay loops, which are installed further along the radiation path. The resulting solution contains up to three subsystems that use fundamentally different element bases: birefringent crystals, block (non-fiber) polarization elements, and fiber-optic elements. Moreover, the last fiber part contains polarization-sensitive elements (fiber circulators and fiber polarization controllers on bending losses). The use of such a “hybrid” pulse sequence formation circuit makes it technically difficult.

US Pat. No. 6,067,311 of May 2, 2003 discloses a method for propagating laser pulses. This method is used to convert a single input laser pulse into four identical output laser pulses, separated in time. An input linearly polarized laser pulse passes through a quarter-wave plate, where its polarization is converted into a circular one. The P and S components of the laser pulse are separated using a polarization divider. The P-component passes through the divider, and the S-component is sent to a delay circuit containing a pair of mirrors and quarter-wave plates. After passing through the delay circuit, the S-component follows the same path as the P-component, but it turns out to be delayed relative to it in time. Further, two time-spaced pulses pass through another delay circuit with a quarter-wave plate installed in front of it. As a result of this, each of the two laser pulses is divided into two more, as a result, a group of four pulses is formed at the output.

The methods described in patents 5293389 from 1994.03.08 and 6067311 from 2000.05.23, are the closest analogues of the developed method of forming a packet of pulses, but they have fundamental disadvantages. The distances between adjacent pulses in the packets obtained by the described methods are determined by the length of the line or delay loop. The delay lines should be of sufficient length so that they can accommodate the elements that control the polarization. This leads to the limitation of the minimum possible distance between adjacent pulses in a packet of several nanoseconds. The total duration of such packets turns out to be quite large, which is why the injection of such a packet into the cavity of a regenerative amplifier of a typical length becomes difficult or technically impossible.

In addition, the method described in patent 5293389, does not allow to obtain a packet of pulses of the same energy. The method in patent 6067311 lacks such a drawback, and can be used to obtain four identical pulses. Also, a fundamental drawback of these methods is their feature, which consists in the fact that the pulses in the generated packets have different polarizations. In principle, all pulses from a packet can be reduced to one polarization, but the possibilities of such a reduction are limited. For example, the speed of high-voltage drivers when turning the polarization of single pulses using Pockels cells limits the minimum allowable distance between pulses and also significantly increases the cost of the system.

Disclosure of invention

Thus, there is a need to create a method devoid of the disadvantages of the closest analogues, i.e. such a method that allows you to generate packets of ultrashort pulses with a distance between adjacent pulses of less than a nanosecond amplified to energies of up to hundreds of microjoules and the following with a frequency of several kilohertz to hundreds of kilohertz.

This problem is achieved in the first object of the present invention with the achievement of the indicated technical result due to the fact that a method for generating ultrashort pulse packets is proposed, which makes it possible to obtain laser pulse packets with a distance between adjacent pulses of less than 1 nanosecond, which consists in repeating the separation of the laser pulse into two pulses, which delayed in time relative to each other and then combined back. Separation and combining of pulses is realized using fiber dividers, and the delay between pulses is due to the fact that the laser pulses between the propagation dividers along the fiber channels of different lengths. The use of several cascades, each of which includes a pair of dividers and fiber channels connecting them, and their combinations allows the construction of circuits forming packets with a different number of pulses in them. It is also allowed to add attenuators to the circuits that attenuate the energy of the pulses propagating in the corresponding channel to change the envelope of the packet being formed.

When used in the composition of laser technological complexes, the fiber design of the packet forming unit according to the first object of the invention allows the simple replacement of one forming unit with another (with other parameters of the formed packages). In this case, laser complexes can be equipped with a set of such blocks, which can be changed by the operator when reconfiguring the complex to other technological processing modes.

A feature of the proposed method according to the object of the present invention is that the distance between adjacent pulses is determined not by the length of the delay line, but by the difference in the lengths of two lines, and therefore it can be reduced to values that depend on the accuracy with which the segments of the optical fibers are measured.

The laser pulse packets formed by one of these methods have a short total duration (for example, with four pulses 150 picoseconds apart, the total duration of the packet is 0.45 nanoseconds), which makes it possible to inject a regenerative amplifier of a typical length (1-2 meters) into the cavity ) using one Pockels cell for a once applied high-voltage pulse (of course, the duration of such a high-voltage pulse should be longer than the duration of the entire packet, but less after the cavity bypass period). After amplification of such a packet in the amplifier, complete unloading of the resonator (i.e., the release of the entire packet of pulses) can also be carried out with one Pockels cell for a single operation.

Brief Description of the Drawings

The present invention is illustrated by drawings in which the same or similar elements are marked with the same reference numerals.

In FIG. 1 shows a general block diagram of a laser complex of the present invention.

In FIG. 2 shows a diagram of a pulser packet forming unit based on fiber dividers.

In FIG. 3a shows a generated burst of pulses.

In FIG. 3b shows a generated packet of equidistant pulses.

In FIG. 4a shows an embodiment of a circuit for a pulse packet forming unit based on fiber dividers with attenuators.

In FIG. 4b shows an embodiment of a circuit for a pulse packet forming unit based on fiber dividers with attenuators forming a packet of three ultrashort laser pulses.

Detailed Description of Embodiments

The present invention is further described with reference to the accompanying drawings by way of examples of its implementation, which are illustrative, but not limiting the scope of the claims of the present invention, defined only by the following claims.

A laser complex that implements the proposed method for generating multi-pulse packets of picosecond / femtosecond laser pulses in the general case contains (Fig. 1): a laser driving ultrashort pulse generator 1, a stretcher 2, a pulse packet forming unit 3, power amplifiers 4, a regenerative amplifier unit 5, decimation unit 6, compressor unit 7, optical harmonic generation unit 8.

A master oscillator with mode 1 synchronization, which in a particular (but preferred) case can be implemented in a fiber version, generates a train of ultrashort pico- or femtosecond laser pulses with a frequency of ƒ ₀ . The train of pulses is fed to the stretcher 2, in which the laser pulses stretch in time (stretching). Preferred is the fiber performance of the stretcher. The stretched pulses enter the pulse packet forming unit 3, where each pulse from the original sequence is converted into a laser pulse packet. Packets of laser pulses at the output of this unit follow each other with a frequency ƒ ₀ . All packages are identical to each other. The laser complex may contain power amplifiers 4 (preferably in fiber design), amplifying laser pulses without changing their repetition rate. Next, the pulses are fed to a laser regenerative amplifier 5. The regenerative amplifier contains all integral functional parts (Faraday isolator, linear or ring resonator, active element (s), pump system, Pockels cells). In a regenerative amplifier, laser pulse packets are amplified based on the principle of laser light amplification by means of stimulated emission. The repetition rate of amplified packets at the output of the regenerative amplifier is reduced to the frequency of operation of the regenerative amplifier itself ƒ ₁ (<ƒ ₀ ), which can range from units of hertz to hundreds of kilohertz. The pump source should provide a gain sufficient to amplify the pulse train to an energy of 0.05 to 0.5 mJ. Such energy at the output of the amplifier is necessary to compensate for the final transmission of the following subsystems of the complex (thinning block 6 and compressor 7), typically 50-70%.

After the regenerative amplifier, the amplified pulse packets pass through a decimation unit 6, which is an optical pulse selector, in which a rectangular or trapezoidal transmission window is formed, covering in time the entire packet of amplified pulses. The decimation unit may be implemented, for example, in the form of an optical circuit with an electro-optical or acousto-optical cell. The decimation unit 6 is capable of operating in an external signal control mode to isolate the required number of laser pulse packets, determined by the specific technological processing mode. Also, the decimation unit can operate in a mode of decreasing the packet repetition rate to ƒ ₂ , expressed as ƒ ₂ = ƒ ₁ / N, where N is a natural number. The amplified packets of pulses are sent to the compressor 7, where they are back-compressed to pico- or femtosecond durations. The laser complex can be equipped with an optical harmonics generation unit 8.

Embodiments of the pulse packet forming unit 3 are described below with reference to FIG. 2, 3a, 3b, 4a and 4b. Each laser pulse 21 enters the first fiber divider 11 through a fiber guide 10, where it is divided into two pulses 22 and 23, each of which propagates along its fiber channel 12 and 13 in the direction of the next fiber divider 14. The fiber channels 13 and 14 have in common case of different lengths L ₁₂ and L ₁₃ respectively. A pair of pulses 22 and 23 arrive at the divider 14 delayed in time relative to each other by the amount δТ ₁ = | α ₁₂ L ₁₂ -α ₁₃ L ₁₃ |, where α ₁₂ and α ₁₃ are the coefficients inversely proportional to the propagation velocity of the pulse along the fibers 12 and 13 Each of these two pulses is split into two more pulses (22 by 24 and 26, 23 by 25 and 27) falling into the fiber channels 15 and 16. The propagation times of the radiation through channels 15 and 16 differ from each other by δТ ₂ = | α ₁₅ -α ₁₅ L ₁₆ L ₁₆ |, where α ₁₅ and α ₁₆ - the coefficients inversely proportional to the velocity of pulse propagation in Loknya 15 and 16, a L ₁₅ and L ₁₆ - respectively, the geometric lengths of these fibers. Each of pulses 24, 25, 26 and 27 is divided by a divider 17 into two pulses, which at the output of the divider 17 are in different channels 18 and 19 (pulse 24 is divided into pulses 28 and 32, 25 - by 29 and 33, 26 - by 30 and 34, 27 - at 31 and 35). One of these formed groups of pulses (28, 29, 30, 31) follows the fiber 18, which is considered an idle channel, and can be used, for example, for diagnostics. Another group of pulses (32, 33, 34, 35) follows the fiber 19, which is considered the working channel, and this group is a formed packet of laser pulses, which will be amplified further.

In order to receive at the output of the pulse packet forming unit 3 four pulses of which no two coincide with each other in time, the lengths of the fiber channels L ₁₂ , L ₁₃ , L ₁₅ , L ₁₆ must be selected so that their pairwise absolute differences of the propagation times they did not coincide with each other, i.e. δT ₁ ≠ δT ₂ . The times δT ₁ and δT ₂ correspond to the distances between pairs of pulses. The generated pulse packet at δT ₁ ≠ δT _{2 is} shown in FIG. 3a.

It is preferable, but not necessary, to use the same fiber in channels 12, 13, 15, 16, while the coefficients α ₁₂ , α ₁₃ , α ₁₅ , α ₁₆ coincide with each other and the delays between pulses introduced by the pairs of channels are determined by the differences in the geometric lengths of the fibers 13, 14, 16, 17 and are δT ₁ = α | L ₁₂ -L ₁₃ | and δT ₂ = α | L ₁₅ -L ₁₆ |, where α is the coefficient inversely proportional to the speed of radiation propagation along the fiber used.

To obtain a packet of four pulses equidistantly spaced from each other, the lengths of the optical fibers are selected so that the value of δT ₂ differs from δT ₁ exactly two times (in any direction, i.e. δT ₂ can be equal to 2δT ₁ or δT _1/2 ) The distance between two adjacent pulses in the packet is equal to a smaller value of δT ₁ and δT ₂ . Formed package of four equidistant pulses? T ₁ =? T _2/2 shown in FIG. 3b.

Since the distance between adjacent pulses is determined by the difference in fiber lengths, this scheme can provide the formation of a packet of pulses with a fairly small distance between adjacent pulses. For example, when using fiber of the Nufern type PM-980-XP at | L ₁₂ -L ₁₃ | = 3 cm, | L ₁₅ -L ₁₆ | = 6 cm, the delay between pulses δT ₁ is 150 ps. The formation by the described method of pulse packets following very short intervals between each other (10 picoseconds or less) is technically difficult, since in this case the interpulse distance becomes comparable with the accuracy with which the fiber sections are measured (10 ps corresponds to the accuracy of measuring the length of the fiber sections approximately ± 0.2 mm). The maximum possible inter-pulse distance is almost unlimited, since fiber channels can have a length of tens to hundreds of meters or more. However, in practice, it is advisable to use packets with an interpulse distance of not more than 10 ns, which corresponds to a fiber length difference of 20 cm.

In FIG. 4a shows a modification of the circuit of the pulse packet forming unit, in which attenuators 36 and 37 are installed in channels 13 and 16, respectively, designed to attenuate laser pulses following these channels. As attenuators, fiber attenuators with bending loss can be used. Using attenuators allows you to adjust the relative amplitude of the pulses propagating in these channels and thereby change the shape of the envelope of the generated packet. Attenuators can be installed in any of the channels 12, 13, 15, 16.

The proposed method for executing a pulse packet forming unit can be modified to obtain packets with a large number of pulses by adding additional serial stages of dividers to the existing circuit with corresponding fiber channels of various lengths. The number of pulses in the generated packets will be 2 ^n-1 , where n is the number of dividers (in the versions shown in Fig. 2 and 4a, n = 3). To form packets with equidistantly delayed pulses, the differences in the propagation times in paired fibers (between two dividers) should be proportional to powers of 2, i.e. δT ₂ = 2δT ₁ , δT ₃ = 4δT ₁ , δT ₄ = 8δT ₁ , etc., where δT ₁ is the minimum difference in propagation times. The sequence of connecting the component cascades can be arbitrary.

The formation of packets with a different number of pulses (not equal to 2 ^n-1 ) is possible due to the use of circuits in which the dividers are not arranged as a sequence of cascades, but in the form of their combinations with each other. For example, in FIG. 4b shows a circuit for receiving a packet of three pulses of the same energy. In this scheme, a packet of three pulses is formed in the output fiber channel 19 of the divider 17. The same energy of all three pulses is achieved using attenuators 38 and 39, tuned to half attenuation.

In the case of the operation of the packet forming unit 3 as part of the laser complex, this unit in the fiber version can be equipped with detachable input-output connections. This will allow for the rapid replacement of one block in the laser complex with another (with other parameters of the packs being formed). In this case, the laser complex can be equipped with a set of blocks 3 with various parameters of the generated pulse packets intended for processing materials in various modes.

Thus, the proposed method allows the formation of packets of ultrashort laser pulses, with intervals between adjacent pulses from 10 ps to 10 ns. The parameters of the resulting pulse packets make it possible to use this method when constructing laser complexes that implement the principle of amplification of chirped pulses to amplify such packets to the energies necessary in the technological problems of microprocessing of various materials.

Claims (14)

1. The method of forming packets of ultrashort laser pulses, which consists in the fact that: - the next laser pulse along the fiber waveguide is divided using the first divider into two pulses; - the said two pulses follow the fiber channels connecting in the second divider and having different lengths, as a result of which the two pulses are delayed relative to each other in time; - each of the two mentioned pulses is divided into two more in the second divider, as a result, two pairs of pulses are formed, which follow two fiber channels connecting in the third divider and having different lengths, as a result of which the two mentioned pairs are delayed relative to each other in time ; - the two mentioned pairs of pulses are fed to the inputs of the third divider, resulting in a sequence (packet) of four laser pulses separated in time on one of the output channels of this third divider. 2. The method of forming sequences according to claim 1, in which an attenuator or attenuators that attenuate laser pulses propagating along this fiber are located in one or more fiber channels located between two adjacent dividers. 3. A femtosecond laser complex containing: - generator of ultrashort laser pulses; - a stretcher for receiving chirped pulses; - a laser pulse packet forming unit operating in one of the ways according to claim 1 or 2; - amplifier or amplifiers of chirped laser pulse packets; - thinning unit, operating in the mode of the pulse selector / pulse packet, at frequencies up to the frequency of the amplifier; - compressor of chirped laser pulses. 4. The complex of claim 4, further comprising an optical harmonic generation unit that provides amplified wavelength conversion.

Images