WO2010089511A1

WO2010089511A1 - Double prism autocorrelation device for the time measurement of ultra-short light pulses

Info

Publication number: WO2010089511A1
Application number: PCT/FR2010/050184
Authority: WO
Inventors: François Etienne Louis SIMON; Ji-Ping Zou
Original assignee: Ecole Polytechnique
Priority date: 2009-02-06
Filing date: 2010-02-04
Publication date: 2010-08-12
Also published as: FR2942036A1

Abstract

The present invention relates to a single-shot optical autocorrelation device for the time measurement of ultra-short light pulses, including: means for dividing an incident light beam at a fundamental optical frequency into two replicated beams; means for recombining said replicated beams, arranged so that the replicated beams cross each other and define an angle other than zero therebetween; a nonlinear optical element placed substantially at the crossing of said replicated beams so as to generate a dual light beam with an optical frequency equal to the sum of the optical frequencies of the two replicated beams; characterised in that the device further comprises, as division means and recombination means, a unitary optical splitter having a shape which is symmetrical to that of the a plane.

Description

- -

"Biprism autocorrelator device for the temporal measurement of ultrashort light pulses"

TECHNICAL FIELD The present invention relates to an autocorrelator device for the temporal measurement of ultrashort light pulses. It also relates to a method for temporally measuring ultrashort light pulses.

The field of the invention is more particularly but in a nonlimiting manner that of the temporal characterization of ultrashort, picosecond, femtosecond and attosecond light pulses.

State of the art

The temporal characterization of ultrashort light pulses, and in particular their shape and duration, is of increasing importance as industrial applications of femtosecond lasers develop. In some of these applications, such as, for example, eye surgery, it is crucial to control both the amount of energy and the instantaneous power delivered by each light pulse, and thus to control both their shape and their stability. . It is therefore also essential to have simple devices to implement and effective to perform these measurements.

In the rest of the sequence, we call "ultrashort light pulses" light pulses whose temporal duration is of the order of one picosecond or shorter (femtosecond or even attosecond). These are pulses whose shape measurements can not be made directly with conventional optoelectronic devices such as photodetectors, slit cameras, etc., because of the frequency bandwidth that would be required. Optical correlators have long been known, which make it possible to measure the intensity correlation function of two light pulses by combining them in a non-linear optical medium so as to generate a signal whose frequency corresponds to the sum of the optical frequencies of the two light pulses. two pulses. When you combine an impulse with your own replica, you end up with an autocorrelator that - - makes it possible precisely to control the characteristics of shape and temporal width of this light pulse. Since the optical frequency of the signal generated in the non-linear optical medium is twice the optical frequency of the pulse, the non-linear mechanisms used in these systems are called "second harmonic generation" or "second generation harmony, SHG "and these systems are called second-order intensity autocorrelators.

In order to be able to describe the totality of the autocorrelation function of the pulse, it is necessary to introduce a variable delay between this light pulse and its replica. This can be done in two ways:

With an optical delay line, in a Michelson interferometer-type assembly, for example. The two replicas of the incident light pulse are separated between the two arms of the interferometer, their relative delay is varied by varying the length of one of the arms of the interferometer, the two pulses are recombined by making them cross in a non-linear crystal and measure the autocorrelation function with a point detector located in the axis of the assembly. Such montages are called multicouple autocorrelators because they require a light pulse by measuring the autocorrelation function, for each delay value. They have the significant disadvantage of being dependent on the repetition frequency of the laser or the amplifier.

It is also possible to generate the optical delay by making two sufficiently large and spatially homogeneous optical beams cross in the non-linear crystal, and recording the autocorrelation function instantaneously with a matrix detector (CCD, photodiode array, etc.). ) placed in the plane of incidence of the two beams. In this case, each pixel of the detector records the autocorrelation function of the light pulse corresponding to a different value of the optical delay. Such mounts are called one-shot autocorrelators because they allow the autocorrelation function to be recorded with a single light pulse. It should be noted, however, that the optical assembly upstream of the nonlinear crystal is substantially the same as for multicouple autocorrelators because It is also necessary to separate and recombine two replicas of the light pulse.

Document WO 02 091116 of Q. Fu, SP Nikitin and AV Masalov, entitled "Appartus et methode de computation intension et phase de une pression d'une interferometric asymmetric single-shot autocorrelator", describes a device based on an autocorrelator assembly. single shot. This device also makes it possible to remove the ambiguity of sign from the time axis, and also to determine the sign of the phase modulation. This result is obtained by combining all or part of the beams from the nonlinear crystal and captured by the CCD detector, and by applying complex algorithms to extract the analytical signal. At the optical level, the separation of the light pulse into two replicas is always done in a conventional manner with a beam splitter. Document WO 02 31456 to N. D. Whitbread and A.C.

Carter, entitled "Optical autocorrelator" which is freed from conventional interferometric optical assembly by realizing the temporal splitting of the light pulse by generating two contrapropagating waves in a waveguide. In this case, the non-linearity is introduced by a two-photon absorption phenomenon in the material constituting the waveguide.

Document WO 04 109345 by G. Ramos-Or-Tiz, M. Cha, SR Marder and B. Kippelen entitled "Third order optical autocorrelator for time domain operation at the telecommunication wavelength" is known which uses a non-linearity of the third order in order to obtain a better sensitivity. This technique requires the introduction of a variable delay between the two replicas of the light pulse, achieved through a conventional optical interferometric optical assembly with two arms before the non-linear element. It also uses an algorithmic complex enough to obtain the shape of the pulse.

These devices make it possible to perform very complete characterizations of ultrashort optical pulses, but they have the disadvantage of requiring the implementation of complex optical and algorithmic means, therefore potentially expensive and delicate. - -

The object of the present invention is instead to provide a device for measuring ultrashort light pulses for routine applications which is instantaneous, simple to implement, compact, robust and of moderate cost. The present invention also provides a method for temporally measuring ultrashort light pulses meeting the same criteria

Presentation of the invention

This object is achieved with a one-shot optical autocorrelator device comprising: means for dividing an incident light beam at a fundamental optical frequency into two replicated beams.

means for recombining said two replicated beams, arranged such that the two replicated beams intersect at a non-zero angle with each other; a nonlinear optical element placed substantially at the point of intersection of said replicated beams, so as to generating a doubled light beam, with an optical frequency equal to the sum of the optical frequencies of the two replicated beams,

Characterized in that it further comprises, as division means and recombination means, a monobloc optical divider having a shape symmetry with respect to a plane.

One-piece optical splitter comprises a rigid element, therefore without moving or adjustable parts, in which the incident beam is separated into two replicated beams by means of refractions and / or reflections, which replicated beams are oriented so that they intersect with a non-zero angle.

Preferably, the two replicated beams are generated by dividing the wavefront of the incident beam.

The monobloc optical divider may advantageously comprise a biprism made of transparent material at a wavelength of interest, of triangular section.

The use of this monobloc optical divider in place of interferometric assemblies based on cubes or blade separators, retroreflectors and mirrors described in the prior art does not constitute a - - simple variant but introduces a drastic simplification of the single-chip optical autocorrelator. Indeed :

The alignment constraints and the mechanical stability requirements of the device, which can be achieved with all the components simply centered on the optical axis of the incident beam, are considerably reduced;

- The need to precisely adjust by adjustment the lengths of the optical paths of the two replicated beams disappears. Insofar as the monobloc optical divider is placed perpendicularly to the incident beam and centered on its optical axis, it generates two replicas of substantially equal intensity, which two replicas cross on the optical axis of the incident beam having traveled substantially the same optical distance.

These provisions make the construction and adjustment of the device particularly easy.

The nonlinear optical element may comprise a quadratic nonlinear crystal cut so as to obtain a phase tuning at the double frequency of the optical frequency of the incident beam. The non-linear crystal may be one of the following materials: BBO, LiO3, KDP, KTP, or any other material to achieve the desired effect at the wavelength of interest.

The nonlinear optical element is preferably of small thickness, for example a thickness between 1 mm and 50 μm, advantageously between 100 μm and 500 μm. The device according to the invention may further comprise optical filtering means placed after the nonlinear optical element and blocking the light at the fundamental optical frequency.

It may also include detection means for measuring the light intensity of the frequency-doubled light beam along a line, a plurality of lines, or a plane. These detection means may advantageously comprise a CCD matrix detector.

The optical filtering means may be made by a difference in sensitivity of the detection means at respectively fundamental and doubled optical frequencies, said detection means being in this case significantly less sensitive to the fundamental optical frequency than to the doubled frequency.

The detection means may be directly contiguous to the output face of the non-linear element. The device may also comprise optical means for imaging the doubled frequency beam of light, with fixed or adjustable magnification.

The device according to the invention may further comprise means for increasing the detection dynamic, inserted in the path of the frequency doubled light beam.

The means for increasing the detection dynamic may comprise a variable attenuation plate inserted into the path of the frequency doubled light beam.

said attenuation blade having different transmission coefficients on either side of its median line,

said attenuation blade being positioned in such a way that the doubled light beam, passing through it, is divided into two half-beams of different intensity.

The two half-beams represent the same function of autocorrelation with different intensities. This takes advantage of the symmetry properties of the generated signal to make the best use of the detection dynamic and in particular to be able to analyze at the same time the overall shape of the pulse and the shape of the curve feet that would otherwise be embedded in the noise. It is important to note that this means of increasing the detection dynamic is not limited to the device according to the invention but can advantageously be implemented in all cases where signals with symmetry properties have to be detected. comparable.

The device according to the invention may further comprise means for replacing during operation the monobloc optical splitter by a diaphragm. Several one-piece optical dividers and diaphragms can be mounted in a moving part. This piece can be for example a turret, a wheel, a translation plate, motorized or manual, moving in rotation or in translation, which positions the - - elements on demand in the incident light beam, at a correct distance from the nonlinear optical element, and aligned with the optical axis.

The device may further comprise alignment means comprising at least two axis markers arranged substantially along the optical axis of the incident light beam so as to materialize the path of the light beam in said device when the divider monobloc optics is removed. The axis markers may include in particular at least two diaphragms, or at least one diaphragm and a numerical mark on the camera. In this way, the device can be easily aligned by positioning it relative to the incident light beam so as to optimize the doubled light beam.

The device according to the invention may further comprise means for calibrating the time transfer function. These calibration means may comprise a blade introducing an optical delay of known value, which can be inserted into the incident light beam before the one-piece optical divider, so as to delay one of the two replicated beams with respect to the other one. a specific quantity. This calibration consists of: measuring the position, for example on the detection means, of the autocorrelation function in the presence and in the absence of the blade respectively, and

- Knowing the delay introduced, calculate the time transfer function of the device from this position difference of the autocorrelation function.

The present invention also relates to a method of single-shot measurement of an optical autocorrelation function according to which:

an incident light beam at a fundamental optical frequency is divided into two replicated beams, said two replicated beams are recombined so that they intersect at a non-zero angle with each other,

a doubled light beam, with an optical frequency substantially equal to the sum of the optical frequencies of the two replicated beams, is generated in a nonlinear optical element placed substantially at the point of intersection of said replicated beams, characterized in that the division and recombination of the beams is performed by a one-piece optical splitter having shape symmetry with respect to a plane.

The temporal resolution of the measurement can be modified by changing the angle of intersection of the two replicated beams, and therefore the size of their interaction zone in the nonlinear optical element. This change of intersection angle of the two beams can be done by changing the monobloc optical divider inserted into the incident beam.

This method gives the device according to the invention a great flexibility of use, allowing the measurement of pulse widths in a very wide range, ranging from picoseconds to attoseconds, while maintaining a simple implementation.

DESCRIPTION OF THE FIGURES AND EMBODIMENTS Other advantages and particularities of the invention will appear on reading the detailed description of implementations and non-limiting embodiments, and the following appended drawings:

FIG. 1 illustrates an embodiment of the device according to the invention,

FIG. 2 is an example of a measurement made with the device according to FIG. 1; FIG. 3 illustrates the calibration procedure of the device according to FIG.

1

FIG. 4 is an example of measurements made during the calibration procedure of the device according to FIG. 1,

FIG. 5 is a diagram of a blade treated in such a way as to present a variable attenuation as a function of the position, making it possible to extend the detection dynamic,

FIG. 6 is an exemplary embodiment of a one-piece optical splitter other than a biprism.

A preferred embodiment of the device according to the invention will be described with reference to FIG.

The incident beam 1 at the fundamental wavelength is typically derived from a pulsed laser. It is reasonable to consider this beam as substantially collimated and spatially homogeneous, which corresponds to the optimal conditions of use of the device according to the invention. A wavefront of this beam is represented in 2. - -

In a prior art optical autocorrelator of the prior art, the incident beam is divided into two replicas and these replicas are recombined by means of an interferometric type assembly based on separate beamsplitters and separate reflection means, which must be adjusted. position and alignment accurately.

In the device according to the invention, these separation and recombination operations are performed by a single piece, the monobloc optical splitter 3. This one-piece optical splitter has a shape symmetry with respect to a plane. It is placed perpendicular to the incident light beam 1, so that the optical axis of said beam is contained in its plane of symmetry. In this way, the monobloc optical divider 3 separates the incident light beam 1 into two replicated beams 4a and 4b which intersect on the optical axis of the incident beam by forming a non-zero angle α between them. According to an advantageous characteristic, the optical path made by the two half-beams is substantially identical. By optical path we mean the geometric distance multiplied by the refractive index of the materials crossed.

According to a preferred embodiment of the device according to the invention, this monobloc optical divider is a biprism of triangular section 3 whose base is of dimensions substantially greater than the diameter of the incident light beam. This biprism is arranged such that the incident light beam penetrates its base perpendicular to the latter, and that the optical axis of said incident light beam passes through its apex. It is advantageously made by machining a block of N-BK7, its base has dimensions of 15 mm x 15 mm and the faces form an angle of 14 degrees with the base. According to another embodiment also implemented in a device according to the invention, the faces form an angle of 24 degrees with the base.

It should be noted that other embodiments of the monobloc optical splitter are possible without departing from the scope of the invention. An example is shown in Figure 6, wherein the incident light beam 1 is divided into two replicated beams 4a and 4b by means of a prismatic element 41 having reflective treatments 42 on parts of its faces. - -

A nonlinear optical element 5 is placed at the intersection of the two replicated beams, so as to generate a nonlinear interaction between these two beams. This interaction can be achieved, for example, by second harmonic generation, autodiffraction, or polarization-enabled second harmonic generation. The nonlinear optical element 5 may for example consist of one of the following materials: BBO, LJ03, KDP, KTP.

Advantageously, this non-linear optical element 5 is a BBO crystal making it possible to implement the second harmonic generation phenomenon, cut and positioned so as to obtain a phase matching at the double frequency of the optical frequency of the beam incident to 1057 nm. Its dimensions are of the order of 15 mm x 15 mm with a thickness of 1 mm. According to another embodiment, its dimensions are of the order of 5 mm × 5 mm with a thickness of 100 μm, or 500 μm, or 50 μm. The angular tolerance allowed with this BBO crystal on the orientation of the polarization of the incident light beam is 22.8 degrees.

The two replicas of the incident beam, crossing each other in the non-linear crystal, give rise, at each point in their overlap zone, to a sum beam 7 of optical frequency equal to twice the optical frequency of the incident beam, whose vector of propagation is the vector sum of propagation vectors of the replicas of the incident beam. This sum beam is called in the following the doubled light beam. This doubled beam 7 is therefore propagated parallel to the optical axis of the device, which can be assimilated in the preferred embodiment to the optical axis of the incident beam.

The X axis is defined as being the axis in the plane of incidence of the replicas, perpendicular to the optical axis of the system. The intensity 8 S (x) of this doubled beam 7 along the X axis is proportional to the value of the intensity correlation function of the two replicas for a temporal shift 2τ, which is proportional to the distance x to l optical axis. This offset is illustrated by the representation of their respective wave fronts 6. With c the speed of light in the vacuum and n the refractive index of the non-linear material we have:

- -

And assuming an intensity of the identical light pulse I (t) in the two replicated beams,

The spatial width of the autocorrelation function, and thus the resolution and measurement range of the device can be adjusted by varying the angle of incidence α of the replicated beams. And as this angle is entirely determined by the monobloc optical divider, just change the latter.

According to a preferred embodiment of the device according to the invention, several monobloc optical dividers 3 are mounted in a wheel, preferably motorized, which allows rotation to change that which is inserted into the beam. These one-piece optical dividers are positioned in the wheel (not shown) so that each of them is placed at the correct distance from the non-linear element so that the replicated beams intersect in the latter. Any other means of replacing monobloc optical dividers, motorized or not, can of course be considered without departing from the scope of the invention.

In this way, the device according to the invention offers a great flexibility of adaptation of the resolution and the measurement range, so that picosecond, femtosecond or even attosecond pulses can be measured under optimal conditions.

According to the preferred embodiment, the device further comprises:

optical filtering means 9, imaging means of the doubled beam 10,

detection means 11.

The optical filtering means are preferably constituted by an optical filtering blade comprising a surface treatment, so as to block the light at the fundamental frequency and let the light pass at the double frequency, said filtering blade being inserted after the nonlinear element.

The optical filtering means 9 may also, without departing from the scope of the invention, consist of: - -

spatial filtering means such as, for example, a diaphragm or a slot, the light at the fundamental optical frequency and the light at the double frequency being separated spatially,

surface treatments applied to all or some of the following elements: exit face of the non-linear element, imaging means, detection means,

detection means sensitive to the doubled optical frequency but significantly less sensitive to the fundamental optical frequency. The imaging means 10 are preferentially but without limitation consisting of one or more LASN9 lenses in particular with focal lengths of the order of 25 mm. They make it possible to image with a fixed or variable magnification the autocorrelation function of the pulse 12 on the detection means 11. In the preferred embodiment, these imaging means allow two fixed magnifications, x1 and x2, selectable by the operator.

The detection means preferably consist of a CCD matrix, advantageously integrated in a 12-bit digital camera. But it may also be non-limiting manner one or more CCD strips or photodiodes, or even a scanning device in front of a photodetector point.

According to another embodiment, the device according to the invention does not comprise filtering means 9, doubled beam imaging 10 and detection 11. The doubled beam comprising the autocorrelation function of the light pulse is directly visible on the exit face of the nonlinear element 5.

According to another embodiment, the device according to the invention does not comprise means for imaging the doubled beam 10. The filtering means 9 and the detection means 11 are directly contiguous to the exit face of the non-existent element. linear 5.

The device according to the invention may further comprise means for increasing the detection dynamic consisting of a variable attenuation plate 13 inserted in the path of the doubled light beam. This attenuation plate 13 has different transmission coefficients on either side of its median line, and it is positioned in the device so that the doubled light beam, passing through it, is divided into two half-beams of different intensity which represent the same function of autocorrelation with different intensities. It is possible in this way to measure the shape of the light pulses with a dynamic intensity much higher than the intrinsic dynamics of the detector, for example to measure both the total height and the shape of the pulse foot of the pulse.

We take advantage of the remarkable property of the device according to which, as illustrated in Figures 2 and 4, the image of the autocorrelation function is visible over a large part of the beam height. Thus, the profiles shown in FIGS. 2 and 4 are only a horizontal line taken in the image of the CCD sensor.

FIG. 5 shows a nonlimiting example of embodiment of an attenuation blade 13. This blade, of N-BK7, of rectangular shape, is separated in the direction of its width into two parts 30 and 31. 31 does not include any surface treatment while the portion 30 comprises an optical density step obtained, according to a preferred embodiment, by a metal deposition of chromium. This optical density step consists of rectangular ranges of uniform optical density, of dimensions greater than the diameter of the doubled beam 7, and whose density values are increasing in the direction 32.

According to a typical but non-limiting example of implementation, the attenuation blade 13 is placed in front of the detector 11 by positioning a suitable density in the beam axis. Two half-images of the autocorrelation function with very different amplitudes are thus obtained simultaneously on the detector, for example:

a half-image where the body of the pulse is saturated but the foot of curve well represented, and another half-image where the pulse is complete and not saturated.

It is possible by moving the attenuation blade, horizontally in the case shown, to adjust the attenuation to the needs. Since the attenuation ratio is known since it is a function of the position of the blade, a profile with a dynamic greater than that of the detector can be obtained numerically by combining the pulse profiles extracted from the two. - - half-images. It is for example possible to obtain with this method pulse profiles with a total dynamic of 10 ⁶ with a CCD sensor whose dynamics is typically a few 10 ³ .

In order to constitute an autonomous device, easy to implement, the device according to the invention may comprise alignment means with the incident light beam: at least two diaphragms 15 are arranged on either side of the monobloc optical splitter , so as to materialize the path of the light beam in the device. According to the preferred embodiment, these diaphragms are largely open during pulse measurements so as not to disturb the beam and can be tightened during the alignment procedure.

During the alignment, the monobloc optical divider 3 is removed from the path of the incident light beam. According to the preferred embodiment, this is done by rotating the wheel carrying the one-piece optical dividers so as to place an empty location in front of the beam. It is also possible to place one of the diaphragms 15 in the constricted position in one of the locations of this wheel. The incident light device or beam is then positioned to optimize the doubled light beam on the detector. Similarly, the device according to the invention may comprise calibration means. This calibration consists in calculating the transfer function of the device, or in other words, in matching a time scale (the duration of the pulse) to the surface of the detector (the number of pixels in the case of a sensor). CCD). This transfer function depends on: the convergence angle α of the two replicas, determined by the one-piece optical divider inserted in the beam, which one-piece optical divider can be changed in the preferred embodiment;

- The growth of the imaging means. According to the preferred embodiment, magnifications of x1 and x2 are possible; - The size and number of points or pixels of the detector.

According to the preferred embodiment, the calibration means are constituted by a plate 14 of known thickness e and of refractive index n _e , as illustrated in FIG. 3. This slide is inserted in front of the monobloc optical splitter 3 so that only the light that constitutes one of the - - replicated half-beams 4a and 4b crosses. A delay is thus introduced between the replicas 20 and 21 of the initial light pulse,

_e = - ck - ^l ) ^e , of known value and which makes it possible to calibrate the complete acquisition chain.

Figure 4 is an example of signals obtained during calibration. Figure 4b is a measurement without the blade 14 while Figure 4a is a measurement with the blade 14 inserted into the incident beam. In the example of FIG. 4, the blade introducing a delay t _e = 1.66 ps generates a positional deviation on the detector of 465-188 = 277 pixels, which gives a transfer function of 1.66 ps / 277 = 6.01 fs per pixel.

FIG. 2 is an example of measuring the width of a light pulse with a device according to the invention. The algorithm used is of course only a non-limiting example to illustrate the operation, but although very simple it remains nonetheless quite exploitable:

the level of the baseline Ymin = 55.6 is detected and the peak of the pulse Ymax = 111;

- We deduce the width at half height of the autocorrelation function Δτ _c of the pulse: X2-X1 = 64 pixels which corresponds to Δτ _c = 384.64 fs; the relationship between the half-height width of the autocorrelation function Δτ _c and that of the pulse Δτ _p depends on its shape, which is known for a given laser. For a gaussian-shaped pulse, as shown in the example of FIG. 2, Δτ _p ≈ Δτ _c / 1.55 = 248.15 fs. Of course, the invention is not limited to the examples that have just been described and many adjustments can be made to these examples without departing from the scope of the invention.

Claims

- -REVENDICATIONS

1. A one-shot optical autocorrelator device comprising:

Means for dividing an incident light beam (1) at a fundamental optical frequency into two replicated beams (4a) and (4b),

Means for recombining said replicated beams arranged in such a way that the replicated beams intersect at a non-zero angle with each other,

A nonlinear optical element (5) placed substantially at the point of intersection of said replicated beams, so as to generate a doubled light beam (7), with an optical frequency equal to the sum of the optical frequencies of the two replicated beams,

Characterized in that it further comprises, as division means and recombination means, a monobloc optical divider (3) having a shape symmetry with respect to a plane.

2. Device according to claim 1, characterized in that the monobloc optical divider (3) comprises a biprism in transparent material at a wavelength of interest, of triangular section.

3. Device according to claim 1 or 2, characterized in that the nonlinear optical element (5) comprises a quadratic nonlinear crystal cut so as to obtain a phase tuning at the double frequency of the optical frequency of the incident beam. .

4. Device according to claim 3, characterized in that the nonlinear crystal (5) consists of one of the following materials: BBO, LJ03, KDP, KTP.

5. Device according to any one of the preceding claims, characterized in that it further comprises optical filtering means placed after the nonlinear optical element (5) and blocking the light at the fundamental optical frequency. - -

6. Device according to claim 5, characterized in that it further comprises detection means (11) for measuring the light intensity of the frequency doubled light beam (7) along a line, a plurality of lines. or a plan.

7. Device according to claim 6, characterized in that the optical filtering means are made by a difference in sensitivity of the detection means (11) at respectively fundamental and doubled optical frequencies, said detection means (11) being significantly less sensitive at the fundamental optical frequency than at the doubled frequency.

8. Device according to claim 6 or 7, characterized in that the detection means (11) comprise a CCD matrix detector.

9. Device according to claim 6, 7 or 8, characterized in that the detection means (11) are directly contiguous to the output face of the non-linear element (5).

10. Device according to any one of claims 1 to 8, characterized in that it further comprises optical imaging means (10) of the doubled frequency beam of light.

11. Device according to any one of the preceding claims, characterized in that it further comprises means for increasing the detection dynamic, inserted into the path of the frequency doubled light beam.

12. Device according to claim 11, characterized in that the means of increasing the detection dynamic comprise a variable attenuation plate (13) inserted into the path of the frequency doubled light beam,

said attenuation blade having different transmission coefficients on either side of its median line, and

being positioned in such a way that the doubled light beam, when passing through it, is divided into two half-beams of different intensity. - -

13. Device according to any one of the preceding claims, characterized in that it further comprises means for replacing during operation the one-piece optical divider (3) by a diaphragm.

14. Device according to claim 13, characterized in that it further comprises several one-piece optical dividers and diaphragms that can be mounted in a moving part.

15. Device according to any one of the preceding claims, characterized in that it further comprises alignment means comprising at least two axis markers (15) arranged substantially along the optical axis of the light beam. incident so as to materialize the path of the light beam in said device when the monobloc optical divider (3) is removed.

Device according to claim 15, characterized in that the axis markings (15) comprise at least two diaphragms.

Device according to claim 15, characterized in that the axis markings (15) comprise at least one diaphragm and a numerical marker on the camera.

18. Device according to any one of the preceding claims, characterized in that it further comprises means for calibrating the time transfer function.

19. Device according to claim 18, characterized in that the calibration means comprise a blade (14) introducing an optical delay of known value, which can be inserted in the incident light beam before the monobloc optical splitter (3), such so as to delay one of the two replicated beams (4a) relative to the other (4b) by a predetermined amount. - -

20. A method of single-shot measurement of an optical autocorrelation function according to which:

an incident light beam at a fundamental optical frequency is divided into two replicated beams (4a) and (4b), said two replicated beams are recombined so that they intersect at a non-zero angle with each other,

a doubled light beam (7), with an optical frequency substantially equal to the sum of the optical frequencies of the two replicated beams, is generated in a nonlinear optical element (5) placed substantially at the point of intersection of said replicated beams, characterized in the division and recombination of the beams is effected by a one-piece optical divider (3) having shape symmetry with respect to a plane.

21. The method of claim 20, characterized in that the temporal resolution of the measurement is changed by changing the intersection angle of the two replicated beams (4a) and (4b).

22. The method of claim 21, characterized in that the change of intersection angle of the two beams is performed by changing the monobloc optical divider (3) inserted into the incident beam.