WORKSHOP ON SCIENTIFIC APPLICATIONS OF COHERENT X-RAYS February 12,1994

Transcription

1 SLAC-R-437 SLAC- 437 SLAC-SSRL 0066 CONF UC-401, 402, 403, , 408, 411 (SSRL-M) WORKSHOP ON SCENTFC APPLCATONS OF COHERENT X-RAYS February 12,1994 ORGANZATON and PROGRAM J. Arthur, G. Materik, H. Winick Stanford Linear Acceerator Center Stanford Synchrotron Radiation Laboratory Stanford University, Stanford CA permanent address HASYLABDESY Notkestrasse 85 D Hamburg, Germany EXECUTVE SUMMARY R.J. Birgeneau-, C.S. Fadey,-* G. Materik **Massachusetts institute of Technoogy Physics, Rm Cambridge MA ***Lawrence Berkeey Laboratory Materias Science Division MS Berkeey CA Prepared for the Department of Energy under contract number DE-AC03-76SFOO515 Printed in the United States of America. Avaiabe from the Nationa Technica nformation Service, U.S. Department of Commerce, 5285 Port Roya Road, Springfied, Virginia 22161

2 Possibiity for one-shot tomography using a high-gain free-eectron aser M. R. Howes*, C. J. Jacobsen**, S. Lindaas** *The Advanced Light Source, Lawrence Berkeey Laboratory, Berkeey, CA USA **Department Physics State University of New York, Stony Brook, NY USA Background Athough x-ray hoography has a ong history, it is ony in recent years that the technoogies needed for it to be successfu have begun to be avaiabe. The advent of unduators and x-ray asers and the prospect of even more advanced pused sources such as high-gain free-eectron asers [Winick have ed to an increase in the eve of interest and activity in x-ray hoography at severa centers around the word. This is based in part on the high interest in three-dimensiona imaging experiments and the reaization that diffraction-tomography experiments in the x-ray region can ony be done via a hoographic type of recording because that is the ony known way to record the x-ray phases. Given that true three-dimensiona imaging wi require a diffraction tomography experiment [Spier, 1987; Devaney; and that x-ray hoography is the right way to record each view [Howes, 19901, we are faced with the chaenge of recording experimentay the many views, probaby dozens, that wi be required. Moreover, it appears as though such a sequence wi not be possibe for unfrozen bioogica materia without ateration of the sampe during data coection. The traditiona response to the damage probem in x-ray imaging is to propose a fash experiment. One of the advantages of hoographic imaging is compatibiity with this type of exposure. However, there is sti a probem in acquiring a sequence of fash images because the first one or first few may sti change the sampe. There are two ways out of the diemma. One is to argue that there is no diemma. Given that the data from each view wi be combined in forming the fina three-dimensiona image, we might argue that the tota counts needed (and hence the tota dose) may not be much greater than for a singe high-quaity two-dimensiona image [Nugent The second way is to perform a fash experiment where a the views are taken simutaneousy [Trebes The atter is the option we want to expore in this presentation. Possibiity for One-shot Fash Tomography - n order to iuminate a sampe with a considerabe number of beams with different directions at the same time we need a device that can receive a singe coherent beam as input and deiver a mutipicity of simiary coherent beams in a variety of directions as output. The setup coud be as 159

3 shown in Fig. 1 with the beam spitter mounted on the upstream side of a thin membrane and the sampe on-the downstream. For the beam spitter to work it woud have to be thin enough to transmit a major fraction of a soft x-ray beam. This suggests a two-dimensiona crysta-ike structure. The advantage of this woud be that the spots in the reciproca attice of a three-dimensiona crysta become rods for a two-dimensiona one. This means that they woud be certain to make a hit with the Ewad sphere and thus woud aways represent an observabe refection.. n other words we woud switch from the Bragg equation to the grating equation. Diffracting structures to reaize the beam-spitter concept Ordinariy, we think of a transmission grating as having a square-wave transmission function with a bar-to-period ratio of about 12. This does what we usuay want which is to concentrate as much diffracted ight as possibe into the + and -1 orders. However we can produce the opposite effect by making the bar-to-period ratio either cose to zero or to one. This means that either the bars woud change into narrow was or the gaps woud change into narrow trenches either of which woud distribute the diffracted intensity over many orders. We can do the same in two dimensions in which case the was or trenches woud be repaced by an array of either towers or hoes. The diffraction. pattern is the same (by Babinet s principe) for both towers and hoes (the sizes being taken as equa) anywhere outside of the footprint of the incident beam. As an exampe, an array of 1Onm hoes on a 50nm grid woud give 4% transmission which woud be divided among 49 beams on a square grid with the orders m, n such that -3 < m < 3, -3 < n < 3 having significant reative strength. f such a system were used to disperse 3nm radiation then a set of 48 hoograms with numerica aperture of /20 (resoution 30nm) and a maximum ange of over 30 between iumination directions, coud in principe be recorded. The ow efficiency of such a system based on hoes in an opaque background can be improved by a factor four by using n-phase-shifting towers on a transparent background or hoes on a K- phase-shifting background. Further detais of projected efficiencies are given in the presentation. Proposed scheme to do a compete tomography experiment using one puse of the high gain free-eectron aser The ayout of the proposed experiment is shown in Fig. 1, to which shoud be added some focusing of the beam and a beam-defining aperture upstream of the sampe. To get a rough idea of the amount of ight we need consider the foowing 160 coherent energy per puse of LCLS = 3mJ fuence needed to expose PMMA at 3OA= 0.5 J/cm2

4 therefore energy needed for a 20x20 km2 hoogram = 2pJ Thus without accounting for any osses we woud have enough energy to make 1500 hoograms. For a usefu reconstruction we woud need at east 30 views so we woud need to maintain an x-ray expoitation efficiency of about 2% which does not seem unreasonabe. Caution Successfu x-ray diffraction tomography experiments wi deiver a vaue for the compex refractive index as a function of position. This quantity is meaningfu because the compex refractive index of a we-defined materias is known and tabuated in the Henke tabes [Henke However if the materias are driven noninear by strong fieds as one gets from a strong pused source then resuting refractive indices which are derived wi not be inteigibe in the usua sense. The x-ray imaging community has not needed to address this matter up ti now and it may turn out that it poses no difficuties at the Stanford LCLS. However the issue shoud be examined in future studies of appications of the LCLS. AcknG+<dgments This research is supported by the foowing United States agencies the NSF under Grant No. DR (SL and CJJ) and Presidentia Facuty Feow Award RCD (CJJ) and The Office of Heath and Environmenta Research of the Department of Energy under Subcontract No A of Grant No. DE-FG02-89ER60858 (MRH). References Devaney, A. J., Reconstructive Tomography with Diffracting Wavefieds, nverse Probems, 2, (1986). Howes, M., Jacobsen, C., Kirz, J., McQuaid, K., Rothman, S., Progress and Prospects in Soft X-ray Hoographic Microscopy, Modern Microscopies Techniques and Appications, Duke, P., Michette, A. G., , Penum, New York (1990). Nugent, K. A., J. Trebes, Signa-to-Noise Ratio and mage Quaity for 3D Mutipe-View X-ray Microscopy, Technica Report, UM-P-93/59, University of Mebourne (1993). 161

5 Spier, E X-ray microhoography-exciting possibiity or impossibe dream, X-ray Microscopy nstrumentation and Bioogica Appications, P. C. Cheng and G. J. Jan, , Springer- Verag, Berin (1987). Trebes, J., J. Brase, J. Gray, R. London, D. Mathews, D. Peters, D. Pinke, G. Stone, M. Rosen, T. Yorkey, X-ray Laser Hoography of bioogica microstructures, X-ray Lasers 1990, G. Taents, , nstitute of Physics Conference Series No. 116, nstitute of Physics, Bristo (1990)..Winick, H., K. Bane, R. Boyce, J. Cobb, G. Loew, P. Morton, H.-D.Nuhn, J. Paterson, P. Pianetta, T. Raubenheimer, J. Seeman, R. Thatchyn, V. Vyet, C. Peegrini, J. Rosenzweig, G. Travish, D. Prosnitz, E. T. Schareman, K. Habach, K.-J. Kim, R. Schueter, M. Xie, R. Bonifacio, L. De Savo, P. Pierini, Short-Waveength FEL s Using the SLAC Linac, US Conference on Synchrotron Radiation nstrumentation, Gaithersburg (1993). Orders 1 Specia diff ratting Resist x-ray detector- Fig. 1. Principes of a possibe method of one-shot fash tomography. A tota of six hoograms are shown but they are part of a two-dimensiona array of 7x r