Multilayer mirrors enable new science using x-ray free electron lasers
The FLASH free-electron laser (FEL) produces intense ultrashort soft x-ray pulses with more than 108 times higher peak brightness than the most advanced synchrotron radiation sources. This allows time-resolved x-ray imaging and holography of nanostructures with a temporal resolution approaching 10fs, which enables new studies of laser-matter interactions and the dynamics of correlated systems. In addition, the ultrafast pulses can be used to obtain structural data before the onset of radiation damage. This vastly increases the usable dose for imaging biological samples and hence improves the resolution of images.
New methods are required to harness the extreme power of the x-ray pulses. The methods developed here will also pave the way to imaging methods for future hard-x-ray FELs. With those sources, atomic-resolution imaging of single uncrystallized macromolecules may become possible.
In the first demonstration of ultrafast x-ray imaging at FLASH, a micron-sized test object was illuminated by a single focused coherent FEL pulse (see Figure 1).1 The coherent diffraction pattern of the object was recorded in the far field on a CCD detector. This pattern was numerically transformed to a high-resolution image of the object, using an iterative phase retrieval technique.2 This image, formed without the use of a lens, has a resolution limited only by the pulse wavelength and the angular extent of the CCD detector. The lensless nature of coherent diffractive imaging offers an advantage: no optical element need be placed near the object nor is it necessary to carefully position the object because focusing is performed numerically in the phase retrieval step.3 However, the experiments at FLASH depended critically on the ability to measure the forward scattering from samples with high sensitivity and low contamination.
The main experimental challenges are posed by the high pulse intensities, which can reach 1015W/cm2 in our experiments. We must prevent the direct (undiffracted) FEL beam from hitting and destroying the direct-detection CCD and to prevent out-of-band radiation (such as plasma emission from the sample) or non-sample scatter from obscuring the coherent diffraction signals. We solved these problems with a unique design that consists of a flat mirror oriented at 45° to the beam with a hole in the middle (see Figure 2). The direct FEL beam passes through a hole in the mirror whereas the diffracted beam is reflected from the mirror onto a bare CCD. Our mirror design enabled the camera to record diffraction angles between −15° and +15°. To reflect scattered light over this wide angular range required a multilayer coating with a very steep lateral gradient. Indeed, the multilayer design had to double in period over only 28mm. Coherent diffractive imaging was performed with cameras operating at 32nm, 16nm, 13.5nm and 4.5nm, each utilizing a different multilayer design.4 The shorter the wavelength the narrower the reflectivity peak width and the higher the specifications for wavelength matching across the optics.
Multilayer coatings are artificial structures that can be designed to enhance reflectivity through constructive interference of beams reflected from the many layer interfaces. Such structures are necessary to efficiently reflect soft x-rays at angles of incidence steeper than the critical angle. At Lawrence Livermore National Laboratory (LLNL), we have lead in the design and fabrication of multilayer x-ray optical components (including lenses, mirrors, beam splitters, and synthetic holographic optical components) for the last 25 years.5 These capabilities were a key ingredient in the success of the Extreme Ultraviolet Lithography project carried out at LLNL and other laboratories.6 Multilayers with nanometer periods are now indispensable in cutting-edge experiments with FELs, not only as x-ray optics but also as samples.
One application used a multilayer film to study the interaction of FEL pulses with matter. In this case, the measurement of the multilayer reflectivity provided a very accurate way to monitor changes in the atomic positions and the refractive indices of the materials in the layers. In experiments at FLASH it was demonstrated that no structural damage larger than 0.3nm occurred within the multilayer during the 25fs FEL pulse.7
The coherent diffractive imaging technique, with its simple experimental design, has produced perhaps the fastest images ever taken.8 In the future we may be able to perform time-resolved imaging of processes induced by FEL pulses of the same or different wavelength, synchronized to the imaging pulse. In this way it will be possible to study ultrafast phase transitions, or dynamic effects such as crack or shock propagation, with nanometer spatial resolution and femtosecond temporal resolution.
We thank Eberhard Spiller and Jennifer Alameda for their valuable contributions in multilayer design and coating. This work was performed under the auspices of the U.S. Department of Energy by the Lawrence Livermore National Laboratory under Contract No.W-7405-ENG-48.
Saša Bajt is a project leader for x-ray optics at Lawrence Livermore National Laboratory. She is currently developing optics for FEL applications and studies of optics damage. She received her PhD from the University of Heidelberg and worked for the University of Chicago developing synchrotron-based x-ray techniques. She received a Hawley medal for the innovation and application of microbeam x-ray absorption fine-structure (XAFS) to mineralogical research. She is an SPIE member.
Henry Chapman is a physicist at Lawrence Livermore National Laboratory, leading a project to develop coherent diffractive imaging at x-ray free-electron laser sources. He developed and carried out the first experiments in diffractive imaging at FLASH and has developed novel time-resolved imaging methods. He received a Bragg medal for the best PhD Thesis in Australia. He is an OSA fellow and a member of SPIE.
