The present invention relates to a method for obtaining a reflective diffraction grating. More particularly, the invention relates to a method making it possible to obtain an optimized dielectric diffraction grating for use under particular conditions.
The invention also relates to the gratings obtained by that obtainment method.
Preferably, but not exclusively, the invention relates to the obtainment of such an optimized grating to perform a high-power laser beam spectral dispersion.
A diffraction grating is an optical device having periodically spaced grooves. It has a diffraction order number that depends on the incident wavelength, the incidence angle, and its period. In the dispersive orders (different from order 0), the reflection angle depends on the wavelength.
Diffraction gratings are used in many optical systems and, in particular, to amplify laser pulses by frequency drift.
Use of Gratings for Frequency Drift Amplification of Pulsed Lasers
Pulsed lasers, or pulse lasers, make it possible to achieve high instantaneous powers for a very short period of time, in the vicinity of several picoseconds (10−12 s) or several femtoseconds (10−15 s). In these lasers, an ultra-short laser pulse is generated by a laser cavity before being amplified in a lasing medium. The laser pulse initially produced, even with low energy, creates a high instantaneous power, since the energy of the pulse is delivered in an extremely short period of time.
To make it possible to increase the power of the pulsed laser without that instantaneous power damaging the lasing medium, it has been considered to stretch the pulse temporally before amplifying it, then to recompress it. The instantaneous powers used in the lasing medium can thus be decreased relative to the power of the pulse ultimately emitted by the pulsed laser. This frequency drift amplification method (often called “CPA” for “Chirped Pulses Amplification”) makes it possible to increase the duration of a pulse by a factor of approximately 103, then to recompress it so that it returns to its initial duration.
This CPA method, described in the article by D. Strickland and G. Mourou, “Compression of amplified chirped optical pulses,” (Opt. Commun. 56, 219-221-1985), uses a spectral decomposition of the pulse, making it possible to impose a path with a different length on the various wavelengths to shift them temporally. The stretching and recompression of the pulses are most often done by dispersion gratings, which have significant dispersive powers and good resistance to the laser flow.
Required Characteristics of These Gratings
The diffraction gratings used to implement this method must meet several particular requirements. They must have a very good reflective efficiency in a dispersive order, i.e., they must reflect a very large proportion of the incident light in a dispersive diffraction order, over a spectral interval corresponding to the spectral interval of the laser pulse to be amplified.
Frequency drift amplification also requires diffraction gratings that have excellent resistance to the laser flow, particularly to recompress a laser pulse after it has been amplified.
Dielectric Gratings
Dielectric gratings, as indicated in the article by M. D. Perry, R. D. Boyd, J. A. Britten, B. W. Shore, C. Shannon and L. Li, “High efficiency multilayer dielectric diffraction gratings” (Opt. Lett. 20, 940-942-1995), have better laser flow resistance performance levels than the more efficient metal gratings. They are made up of a stack of thin dielectric layers placed on a substrate and reflecting up to approximately 99% of the incident light. The upper surface is periodically etched to as to obtain the diffraction grating.
The thicknesses of each of the layers of this stack are chosen so as to form a Bragg mirror, or “quarter wave mirror,” in which layers with a high refractive index nH are alternated with layers with a low refractive index nL. The thicknesses tH and tL, respectively, of the high refractive index layers nH and the lower refractive index nL are determined by the following relationships:
in which:
λ is the wavelength of the incident light;
θH and θL are calculated by the following relationships:
in which θi is the incidence angle of the light on the grating. Such a Bragg mirror makes it possible to reflect, owing to constructive interference phenomena, up to more than 99% of the incident energy for a given wavelength.
However, since the thicknesses of the different layers are calculated for a single wavelength λ, they do not make it possible to obtain satisfactory results for pulses having a spectral width larger than approximately 20 nm, centered on that wavelength.
Drawbacks of the Prior Art
These dielectric gratings based on Bragg mirrors, which are satisfactory for the frequency drift amplification of laser pulses with a spectral width in the vicinity of several nanometers, are not adapted to the shortest pulses, which have a larger spectral width.
To decrease the duration of the pulses, it therefore becomes necessary to have diffraction gratings having optimal performance levels over a wide spectral band of several tens, or even several hundreds, of nanometers. No diffraction grating of the prior art guarantees good performance levels over such a spectral width and a high damage threshold.
The present invention aims to offset these drawbacks of the prior art.
Thus, the invention aims to provide a method making it possible to obtain an optimized dispersive reflective diffraction grating for a particular use.
In particular, the invention aims to make it possible to obtain an optimized diffraction grating for use over a frequency range several tens, or even several hundreds, of nanometers wide.
The invention particularly aims to make it possible to obtain such an optimized diffraction grating for frequency drift amplification of an ultra-short pulse laser having a spectral width of several hundred nanometers and good resistance to the laser flow.
These aims, as well as others that will appear more clearly hereinafter, are achieved by a method for obtaining a reflective diffraction grating for the diffraction of a light beam with a predetermined spectral range, incidence angle, and polarization, including a stack of at least four planar dielectric material layers, an upper dielectric material layer being etched so as to form a diffraction grating, the etching period of which is predetermined.
This method according to the invention implements the following steps:
Preferably, the non-etched layers of dielectric material are placed on a metal layer, and there are between 5 and 15 of them.
Advantageously, the etching parameters whereof the value varies during the computation step are the etching depth and the groove width.
Advantageously, the digital computation of the reflection and/or transmission efficiencies of at least one of the diffraction orders is done for a sample of at least 10 frequencies distributed in a spectral range with a width larger than 100 nm.
According to one preferred embodiment, this spectral range is between 700 and 900 nm.
The present invention also relates to a reflective diffraction grating including:
Such a diffraction grating is therefore different from those based on a Bragg mirror, in which all of the layers of a same index have the same thickness.
Preferably, this reflective diffraction grating comprises at least two layers of silica (SiO2) and two layers of hafnium dioxide (HfO2), alternating, and the etched upper layer is made from silica (SiO2).
Advantageously, such a reflective diffraction grating, for the diffraction of a light ray with a spectral range between 700 and 900 nm, having an incidence angle between 50° and 56°, comprises a substrate on which at least the following are deposited:
According to one advantageous embodiment, such a reflective diffraction grating comprises a layer of alumina deposited between the last layer of hafnium dioxide (HfO2) and the layer of etched silica (SiO2).
The invention also relates to a reflective diffraction grating, comprising a substrate on which the following are successively deposited:
Other aims, advantages and features of the invention will appear more clearly in the following description of one preferred embodiment, which is not limiting on the subject-matter and scope of the present patent application, accompanied by drawings, in which:
Reminder of the Prior Art
Dielectric gratings with too many layers present cracking risks when they are exposed to laser flows. To avoid this drawback, a layer of gold (not shown) can be inserted between the glass substrate 13 and the dielectric stack forming a Bragg mirror so as to reduce the number of thin layers needed to obtain a high reflectivity, while guaranteeing a damage threshold close to those obtained with completely dielectric mirrors.
In that case, the thickness of this layer of gold is much larger than the skin thickness, typically 150 nm, such that the glass substrate has no optical interaction with the laser pulse.
The number of dielectric layers above the gold deposit can be set by the user but, contrary to completely dielectric depositions, it is possible to reduce it to six. This solution is described in the article by N. Bonod and J. Neauport, “Optical performances and laser induced damage threshold improvement of diffraction gratings used as compressors in ultra high intensity lasers” (Opt. Commun., Vol. 260, Issue 2, 649-655-2006).
The upper layer 15 is etched to form the grating. The period and the etching geometry are defined so as to collect the greatest portion of the incident energy reflected in the dispersive diffraction order (−1). Only the energy collected in this diffraction order (−1) will be used in the final laser pulse. The energy emitted in the other orders is lost. The period and the etching geometry are generally defined so as to collect approximately 95% of the incident energy reflected in the diffraction order (−1).
Such a grating of the prior art can only offer good performances for a given wavelength, and is in particular not adapted to the dispersion of a laser pulse covering a wide frequency range.
Sizing Methodology
The present invention is based on the joint optimization of the thickness of the planar layers and the etching profile of the grating. The thicknesses of the different layers are therefore not those determined for the Bragg mirrors, but are each optimized, in connection with the characteristics of the etching profile, by a digital optimization method, to have good reflected efficiencies over a wide spectral width.
The grating to be optimized has a certain number of parameters that are chosen before implementing the optimization method. These parameters are primarily:
The optimization is done by choosing the best combination of values for the following variables:
For each of these values, a minimum and a maximum are determined, as well as an incrementation pitch The minimum and maximum can be chosen in particular as a function of the manufacturing constraints. The incrementation pitch is chosen as a function of the precision of the desired optimization. Furthermore, the incrementation pitch and the [minimum; maximum] intervals are chosen as a function of the computation power available to perform the optimization. The number of computations in fact increases when the intervals are increased or when the incrementation pitches are decreased.
The diffraction grating having these parameters can be dimensioned, according to the invention, with the method comprising the following steps:
A plurality of possible configurations of the diffraction grating are determined corresponding to the aforementioned parameters. To that end, a computer is used to determine all possible combinations by varying the thicknesses of each of the layers of dielectric material and the etching parameters of the upper layer within predetermined intervals and according to the predetermined pitches.
For each of the configurations determined in the first step, the reflected efficiency is computed in the diffraction order (−1) of the grating, for a sample of frequencies chosen in the spectral range of use for the grating to be dimensioned.
After computing the efficiency of each of the configurations, the configuration(s) whereof the efficiencies and characteristics best correspond to the anticipated use of the diffraction grating are selected, using a suitable criterion.
It should be noted that the values of some of the variables can be set, to simplify the computations or if it is not relevant to optimize them. Thus, for example, it is possible to set the thickness of a dielectric layer that does not have a substantial optical effect, such as a fine layer of alumina (Al2O3) present to meet mechanical constraints. The optimization according to the invention can only, however, be done by simultaneously optimizing at least one of the etching parameters (etching height h, incline angle α of the trapeziums, width c of the etched groove) and the thickness of each of the dielectric layers having a significant optical effect, of which there are at least four.
In a novel manner, this digital optimization method therefore takes into account both the thicknesses of each of the layers foaming the grating, and the etching characteristics of that grating.
To determine the plurality of possible configurations, in the case where there are six layers of dielectric materials in addition to the etched layer, software is used that will use the following variables:
The following parameters are entered into the software:
The software initializes each of the variables h, e1, e2, e3, e4, e5, e6, and c at their respective minimum values hmin, e1min, e2min, e3min, e4min, e5min, e6min, and cmin. The reflected efficiency of this first configuration is then computed using the appropriate method for resolving the Maxwell equations.
The first parameter h is incremented by the value of the pitch Δh, while its value is less than or equal to hmax. For each of the values assumed by h, the reflected efficiency of the corresponding configuration is computed using the appropriate method for resolving the Maxwell equations.
The second parameter e1 is incremented by the value of the pitch Δe1, while its value is less than or equal to e1max. For each of the values assumed by e1, the value of h is varied as described above and the reflected efficiency of all of the corresponding configurations is computed using the appropriate method for resolving the Maxwell equations.
The third parameter, then each of the following parameters, is thus incremented until the reflected efficiencies of all of the possible grating configurations whereof the parameters h, e1 , e2, e3, e4, e5, e6, and c are between the set minimum and maximum values, with the set incrementation pitches, have been computed.
Thus, if the following parameters are entered:
Computation of the Reflected Efficiency
For each of these configurations, the reflected efficiency of the grating can be computed for several previously-selected wavelengths, distributed in a given frequency range.
The method for computing the reflected efficiency in the diffraction order (−1) of the configuration of each configuration of the grating, based on a rigorous resolution of the Maxwell equations, rests on the development of the electric and magnetic fields in a Fourier series, which makes it possible to reduce the Maxwell equations to a system of differential equations of the 1st order. Integrating this system of the substrate into the superstrate makes it possible to precisely compute the reflection and transmission efficiencies of the periodic component. A second integration makes it possible to reconstruct the electromagnetic field in the entire space.
This computation method is fully described in the work by M. Nevière and E. Popov, entitled “Light propagation in periodic medias; differential theory and design” (Marcel Dekker, New York, Basel, Hong Kong, 2003).
Once this reflection calculation in the −1 order is done for all of the configurations, it is possible to choose the configuration(s) having both good reflected efficiencies and characteristics compatible with the anticipated use of the diffraction grating.
Parameters Chosen to Obtain the Grating of
The diffraction grating shown in
The incline angle α of the trapeziums forming the etching is chosen at 83°. This angle is closest to the angles measured on the gratings currently made by manufacturers in this type of oxide, and for this type of depth.
It has been chosen to manufacture this grating with three planar layers 21, 23, and 25 of SiO2, alternating with three planar layers 22, 24, and 26 of HfO2, the lower layer 21 of HfO2 being placed on a layer of gold 20.
For each planar layer 21, 23, or 25 of SiO2, the chosen incrementation pitch is 10 nm in an interval of [100; 400] nm.
For each planar layer 22, 24 and 26 of HfO2, the chosen incrementation pitch is 10 nm in an interval of [0; 300] nm.
An additional upper layer 28 of SiO2 is etched over the entire height thereof.
A layer 27 of Al2O3 with a thickness of 50 nm is provided between the upper layer 28 of SiO2 intended to be etched and the upper layer 26 of HfO2 to facilitate the etching of the layer 28 of SiO2 over the entire thickness thereof without damaging the layer 26 of HfO2. This fine layer 27, when it is indispensable to produce the grating, is taken into account in the computations of the reflected efficiency of the grating as a constant. This layer of Al2O3 could, of course, not be used, or could be placed in another position, in other embodiments of the invention.
The interval chosen for the c/d parameter is [0.55; 0.75], with an incrementation pitch of 0.1.
The interval chosen for the etching depth h (which, in this embodiment, corresponds to the thickness of the etched layer) is [300; 800] nm, with an incrementation pitch of 10 nm.
The reflected efficiency in the order −1 is computed for 41 wavelengths comprised between 700 nm and 900 nm.
As a function of the chosen parameters, the number of computations of the reflected efficiency of the different possible configurations of the diffraction grating is therefore 41*3*51*[31]n, where n is the number of planar layers, or 6.
It should be noted that the number of wavelengths for which the reflected efficiency in the order −1 can rise to several hundred for a fine optimization.
Optimization of the Grating Parameters
The computation of the reflected efficiency in order −1 of all of these configurations is done by computer, using the computation method described above.
This method can of course be used iteratively. Thus, when a first implementation of the method makes it possible to detect optimized grating solutions, one or more new implementations with differently chosen intervals and reduced incrementation pitches make it possible to precisely define the best grating solutions.
Using the sizing method according to the invention thus makes it possible to find different grating configurations, having the parameters described above relative to
One of these configurations corresponds to a grating made up of a glass substrate, on which are successively deposited:
The etching is done so that the value of c/d is equal to 0.65.
The etching parameters have been chosen so that the number of diffraction orders is limited to two (order −1 and order 0) so as to limit the distribution of the energy in too many orders. The order 0 not being dispersive (the diffraction angle in that order does not depend on the frequency), the order (−1) in which the incident light is dispersed.
The graph of
The grating sized using this method can then be manufactured by using the traditional manufacturing methods, known by those skilled in the art to manufacture gratings based on Bragg mirrors.
Intervals Allowing the Best Reflected Efficiencies
By using this sizing method, it is possible to determine intervals in which the thicknesses of the layers of a grating having six layers of SiO2 and HfO2 in addition to the etched layer must be located so that the reflected efficiency average in the order −1 of a laser pulse, for example amplified by a material of the Titanium-Sapphire type, with a spectral width of approximately 200 nm centered on 800 nm, arriving on the grating with an incidence comprised between 50° and 56°, is greater than 90%.
The etching depth of this grating is comprised between 625 nm and 775 nm, and the number of lines per mm is comprised between 1400 and 1550.
The intervals in which the thicknesses of the layers are comprised are:
Using a grating having these features is therefore particularly advantageous, in particular to compress a laser pulse amplified by a material of the Titanium-Sapphire type.
Number | Date | Country | Kind |
---|---|---|---|
09 59 157 | Dec 2009 | FR | national |
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/FR2010/052684 | 12/13/2010 | WO | 00 | 8/16/2012 |