XUV optical components (spectral range 3-60 nm) are of great interest for the scientific community, due to the development of new high brilliance sources (X ray laser, high harmonic generation sources, free electron laser), and of new applications (EUV lithography, dense plasma diagnostic, attosecond science, solar imaging telescopes...). In this spectral range multilayer structures are the only possibility to make near normal incidence mirrors. To design and to make such components it is necessary to know perfectly the dispersion of index for the materials used. Because of the strong absorptions in this range of wavelengths index measurements are critical and yet experimentally difficult. The different index values reported in the literature are not in agreement. It is essential to determine the index of materials in XUV range by a systematic and rigorous method. In this context, we propose an innovative project based on XUV interferometry measurements to have direct access to thin film optical constants. These reliable values will be used in this project for the design, optimization and realization of new kind of XUV multilayer components. We decided to focus our efforts on one major evolution of multilayer mirrors, related to the generation of ultrashort XUV pulses: the design and realization of atto-mirrors (i.e. mirrors for attosecond pulses). The phase matching of a XUV harmonic broad spectrum can be achieved with high efficiency by using such mirrors. The need for shorter pulses is explained by the desire to study various ultrafast physical processes (for example fast molecular, solid state and atomic phenomena). The aim of this project is to design and study theoretically and experimentally such multilayer atto-mirrors, by using reliable optical constant values measured by XUV interferometry. That requires also optical and physico-chemical characterization of thin film materials. Interferometric system allows a direct determination of the index of refraction. Such tool requires the knowledge to design very great quality instrument, because of the narrow tolerances in this spectral range (mechanical assembly and polishing of optic). A wave front division interferometer was recently demonstrated and our project will be strongly based on this unique ability of the Institut d'Optique. In order to optimize these atto-mirrors we will study the influence of the deposition conditions on the indices and on their optical properties. Comparisons between measurements and simulations using the measured index will allow a better understanding of component defects (interface, roughness, purity) and of their influence on atto-mirror optical properties. Institut d'Optique has acquired, since more than 20 years, a know-how in the design and realization of new XUV multilayer mirrors and also in the development of XUV interferometers. Numerous collaborations in the past or in progress (NASA, ESA, LIXAM, JAERI, LLC, CEA…) show a great involvement of the Institut d'Optique in this very effervescent domain. The Laboratoire de Chimie Physique – Matière et Rayonnement plays also a major role in in this field of research and will bring to the project its skills in term of optical and physico-chemical analysis of interfaces and multilayers. Based on the synergy of various skills, this project creates a new research topic at Institut d'Optique, 'optical metrology in the XUV domain', and leads to the realization of new optimized XUV components. The expected results will be very useful for the XUV community (solar imaging, plasma diagnostic, attosecond optic) and will strongly stimulate the activities of the partner teams in this field.

Except the synchrotron radiation or the free electron laser, which are only available in large facilities, there is at present no tunable source of extreme ultraviolet radiation (EUV). Yet the use of EUV radiation is quickly developing, mainly because of the implementation of the EUV lithography dedicated to the manufacturing of the next generation of electronic chips and also to research in solar and extra-solar imaging from spatial telescopes and “attosecond” physics. The only commercially available EUV sources are based on the emission of discrete lines from plasmas obtained under various conditions (discharge in gas, electron cyclotron resonance). Nevertheless an energy tunable and quasi-monochromatic EUV source would be of considerable interest to characterize optical components and detectors in this spectral domain. The interaction of relativistic or non-relativistic electrons with solid radiators appears as a very promising technique to realize such a source. Two phenomena have given rise to advanced studies: the transition radiation (TR) and the parametric radiation (PR). The TR occurs when electrons cross the interface between two materials; sources of X-rays based on TR and using periodic multilayer structure as radiators were proposed; their feasibility was experimentally evidenced. The PR is emitted when electrons go through periodic media (crystal, multilayer interferential mirror) close to the Bragg conditions. Once more, X-ray sources based on this mechanism were proposed, built and successfully tested. The main drawback of these tested sources is the need for relativistic electron beams supplied by expensive accelerators (LINAC, storage ring, betatron). Our TPLUS (Tunable Parametric Laboratory UV Source) proposal of tunable source calls upon non relativistic electrons supplied by a 100 kV gun producing PR in a multilayer nanostructure. The main features of such a parametric UV source were studied by a French team in the case of medium relativistic particles and recently by a Russian-American collaboration in the case of non-relativistic particles. Their calculations demonstrate that PR of wavelength ? = 36.4 nm (E = 34 eV) is produced upon irradiation of a Sc/Al multilayer target by 100 keV energy electrons under a glancing angle equal to 43°. Furthermore it is possible to get a yield of approximately 10-5 photon per electron. With an electron gun providing 1 mA, theory estimates an intensity around 109 photons/s/eV/sr (7.108 photons/s/sr for a spectral bandwidth equal to 0.7 eV at 34 eV); this value is lower than the 1015 photons/s/sr for both He I and He II lines supplied by a He-plasma, generated with the ECR technique. Nevertheless this He source can only emit a few intense discrete lines. To our knowledge, no source similar to the one described in this proposal has been built and characterized. Nevertheless different institutional teams (in particular in Russia) and research teams from private firms in USA (Intel Corp., Adelphi Technology) have shown a great interest for this kind of source through several publications and communications to conferences (for instance RREPS-07). The main objective of this proposal is to design and built a tunable EUV laboratory source, to measure its performances and finally to obtain a pre-industrial prototype. This source will be used in the future to characterize EUV optics at their wavelength of application and should allow performing reflectivity measurements of mirrors in the whole EUV spectrum. Indeed, to explore the whole EUV range, specific materials are required to fabricate optimal multilayer mirrors. The data of the PXRMS (Physics of X-Ray Multilayer Structures) survey show that numerous multilayers are designed for applications ranging throughout the EUV range. Thus it is important to have a laboratory tunable source to get some independence towards the allocated synchrotron beam times and also to prepare in an extensive way the synchrotron experiments.