Recherche

Laboratoire Charles Fabry has been involved for a long time in the study and development of EUV telescopes for the observation of the solar corona, in collaboration with Institut d’Astrophysique Spatiale, Orsay. Following the success of the SoHO (1995) and STEREO (2006) missions, we are supported by the French Space Agency (CNES) to participate in the European consortium on the EUV imagers (EUI consortium, PI= Centre Spatial de Liège) for the European Space Agency Solar Orbiter mission. During the last five years, we have provided the optics and multilayer coatings for the two EUV telescopes, a single-mirror telescope that will give a full image of the Sun (FSI, Full Sun Imager) and a two-mirror telescope giving a detailed image of a portion of the Sun (HRI, High Resolution Imager). Spherical surfaces with a very high precision in shape (< 2 nm RMS) and a very low roughness (< 0.2 nm RMS) were first fabricated by our optical workshop with the classical polishing techniques. Then we transformed these spherical surfaces into hyperbola, while preserving the roughness by using a home-developed broad ion beam etching, both in axi-symmetrical and off-axis versions. Concerning EUV multilayer coatings, we have successfully deposited Al-based multilayers, developed recently in our lab, which provide record reflectivity (>55% at 17.4 nm). For the FSI telescope, we have proposed a new concept of dual-band EUV mirrors, made by stacking 2 periodic multilayers spaced by a thin layer and which reflect efficiently two spectral lines with enhanced spectral selectivity. All the fabricated mirrors and coatings were within the demanding specifications of the instrument and represent state-of-the-art EUV optics for space missions. 

Developing imaging methods operating at resolutions outperforming that of optical microscopes is of paramount importance in biology. Of particular interest is X-ray microscopy in the water-window (between Oxygen and Carbon K-α edges, corresponding to a wavelength between 2.4 and 4.4 nm), since it enables biological imaging with endogenous absorption contrast and high resolution thanks to the short wavelength. In this context and as part of the Morphoscope project (Equipex 2012-2020), we develop in collaboration with Soleil Synchrotron, a full field, near normal incidence, mirror-based microscope at a wavelength near 3.14 nm. This development involves addressing multiple technological challenges in term of optical surfaces, coating, mechanics and metrology. The optical design is based on a Schwarzschild objective (two concentric spherical mirrors) with 0.15 numerical aperture and with a long working distance so that the installation of the sample will be facilitated. Several studies of multilayer coating with sub-nanometer thick layers have been carried out in order to obtain the highest reflectance possible near normal incidence at a wavelength near 3.14 nm. A peak reflectance of 23% for an incidence angle lower than 5° has been measured on the metrology beamline at synchrotron Soleil. These results are the best obtained in this spectral range [Burcklen et al. Optics Letters, 42,1927 (2017)]. A strong effort will also have to be made on the lateral gradient of period necessary on the secondary mirror due to the relative high numerical aperture. In order to perform the metrology on the coated mirrors a new five-circle reflectometer is currently developed at Soleil. The silica super polished mirror blanks was performed by our optical workshop and controlled in our laboratory with a high precision phase-shifting Fizeau interferometer. The shape defects are of the order of 1 nm RMS. The roughness is less than 0.1 nm. These very small defects then make it possible to obtain a resolution better than 300 nm. The fabrication of the mechanical parts is in progress. We are considering final assembly and preliminary testing in 2019.

In addition to the multilayer coatings developed for the EUV telescopes and soft x-ray microscope described above (see section A1), we studied and optimized coatings for innovative instruments: x-ray monochromators based on the concept of alternate multilayer grating (section A2-1) and broadband spectrometer for plasma diagnostics (section A2-2). Moreover, we started a more fundamental research on the determination of optical constants in the EUV/soft x-ray spectral domain (section A2-3).

X-ray imaging systems and X-ray spectrometers are key diagnostics for Inertial Confinement Fusion (ICF) experiments. In close collaboration with CEA, we are developing advanced X-ray diagnostics to probe dense plasmas produced in the future Laser MegaJoule (LMJ) facility. Concerning broadband spectrometry diagnostics, there was a specific need to develop multilayer optics with a specified (non flat) spectral response and a good rejection of unwanted bands (at both higher and lower energies). We have successfully designed and fabricated several non-periodic multilayer mirrors in order to reflect photons with a pre-defined reflectivity profile in a specified energy range between 2 keV and 6 keV [Emprin et al., Optics Express, 22, 25853 (2014)]. 100-layer Cr/Sc and 80-layer Ni/SiC with W barrier layers have been optimized and coated on diagnostic prototypes that are now installed on the LMJ facility [Troussel et al., Review of Scientific Instruments 85, 013503 (2014)]. These results provide the first experimental evidence of apodization and spectral shaping in the soft X-ray domain and open the way to more complex optical functions.

In collaboration with LBNL and LLNL, we have measured of the complex index of refraction of sputtered Chromium thin films for photon energies ranging from 25 eV to 813 eV, including the first highly resolved photoabsorption coefficient values in the region of the L2,3 edge [Delmotte et al., accepted in Journal of Applied Physics]. Chromium is a material of wide interest for science and technology in the soft X-rays and extreme ultraviolet (EUV) spectral range and these new data will be of considerable interest for a large community of physicists. The newly determined Cr optical constants will be disseminated in the scientific community and will enable simulations of Cr-based optics in the EUV and soft x-ray range with significantly improved accuracy. 

For the past thirteen years, we have been developing multilayer coatings between 30 eV to 90 eV for the post-compression of chirped attosecond pulses obtained through high order harmonic generation (HHG) [Diveki et al., Journal of Modern Optics 61,122 (2014)]. It requires a spectral phase engineering of the multilayer structures. The measurement of photocurrent at the top of the structure was used to obtain a full phase characterisation of such mirrors on a synchrotron beamline. In this framework we desired to extend this approach at higher energies, especially around 400 eV where a strong effort is currently done to obtain efficiently HHG sources. In this spectral range, the effects of the mean free path of the electrons inside the stack can no longer be neglected, which prevents the phase reconstruction by conventional photocurrent measurements. We have developed a new model taking into account this phenomenon. This approach has been validated through a numerical and experimental study of Chromium/Scandium multilayers used near 360 eV. This work constitutes the first measurement of the phase of a multilayer mirror in the soft X-ray range [de Rossi et al., Optics Letters 40, 4412 (2015)]. 

Conventional synchrotrons emit XUV light pulses in the picosecond range, which is usually too long to study the probed phenomena at their natural timescale. The past 20 years have seen the development of ultrashort XUV sources, emitting pulses with femto or attosecond duration. Especially attosecond sources based on high-order harmonic generation now enable one to revisit synchrotron experiments (photoemission spectroscopy, absorption spectroscopy…) with the suitable time resolution. These experiments require well-characterized XUV pulses to trigger the ultrafast phenomena. In addition the pulses must be highly coherent, that is to say that these waveforms must remain as identical as possible throughout the experiment to enable a clear result to emerge from the measurement process. There is therefore a major need for experimental techniques giving access to the temporal properties as well as the state of coherence of these XUV pulses. Although there exists well-established attosecond temporal measurement techniques, all of them assume the full coherence of the pulse under test. In 2015, we reinterpreted these historical techniques in attosecond physics and showed how to extract the coherence information from existing measurements, giving access to a much richer physics at almost no experimental cost [Bourassin and Couprie,Nature Communications 6, 6465 (2015)]. We established collaboration with the attophysics group at CEA Lidyl, a group at the forefront in this research field, to implement this novel approach. We measured a degree of coherence of the attosecond pulses of 11% when the experiment was operated in routine conditions [Bourassin et al., submitted to Nature Physics], which is far from the generally assumed full coherence.