M e a s u r e m e n t o f X - r a y P u l s e W i d t h s b y I n t e n s i t y I n t e r f e r o m e t r y Ultrafast X-ray pulses provide a powerful probe for investigating structural dynamics in biologica l and material sciences. T he upcoming linac-based undulator s ources are c apable of generatin g brilliant X -ray pulses of ~ 100 femtoseconds (fs). Although measurement of such ultrafas t pulse width is crucial, no methods applicable to hard X- rays are currently available. In this report, we show that intensity interferometry, which is a technique initially developed by Hanbury-Brown and Twiss [1 ] and recently extended to the X-ray region [2,3], is capable of measuring X-ray pulse width (32 ps in FWHM) at SPring-8 [4]. Notably, t he method ca n be easily extended to fs r egion by using simple r monochromators . Most hard X -ray s ources t hat are either currently available or under development , including self-amplified s pontaneous emission ( SASE) free electron laser (FEL) in t he linear regime, are Fig. 1. Schematic view of the experimental setup. Undulator radiation is premonochromatized with a Si 111 double crystal monochromator (DCM). The 4-bounced monochromator and the 4-quadrant slit were employed to extract the longitudinally and transversely coherent portion of the beam, respectively. Two semi-transparent avalanche photo diodes (APDs) were aligned in tandem on the light axis. Outputs of the detectors were connected to the coincidence circuit. considered to generate chaotic light. In this case , intensity interference is observed as an enhancement of the coincidence rate between the two detectors that receive the spatially and temporally coherent portions of t he beam. In particular, w hen th e method is applied to pulsed beam, the enhanced ratio i ncludes information on t he temporal pulse width s t with respect to the longitudinal coherenc e time σ t . Because the coherence time σ t is directl y given by t he energy bandwidth ∆ E of light, evaluatio n of t he enhanced ratio with k nowledge of th e bandwidth enables to determine the pulse width s t . Experiments were performed using beamline BL19LXU , which is equipped with a 27-mete r undulator, the most brilliant X-ray source currently available. T he experimental setup is s hown in Fig. 1 . A monochromator consisting of 4-bounced asymmetric reflections (horizontal diffractions of Si 11 5 3, asymmetric angles α = 78.4º) was used as 107 ⊗ 4 - Q u a d r a n t s l i t 4 - B o u n c e d m o n o c h r o m a t o r A P D s 1 & 2 C o i n c i d e n c e c i r c u i t D C M 2 7 m U n d u l a t o r Fig. 2. The longitudinal mode number M t vs . the energy bandwidth ∆ E. The line shows the fit result with a pulse width st of 32.7 ps . Fig. 2 Makina Y abashi SPring-8 / JASR I E-mail: yabashi @s pring8.or.jp References [1] R. H anbury-Brown and R. Q. Twiss, Nature (London) 177 (1956) 27. [2] E. Ikonen, Phys. Rev. Lett. 68 (1992) 2759; Y. Kunimune et al . , J. Synchrotron Rad. 4 (1997) 199; E. Gluski n et al . , ibid . 6 (1999) 1065. [3] M. Yabashi et al. , Phys. Rev. Lett. 87 ( 2001) 140801. [4] M. Y abashi, K. Tamasaku and T. Ishikawa, Phys. Rev. Lett. 88 (2002) 244801. [5] H. Ohkuma et al . , Proc. of t he 2001 Particle Accelerator Conference, Chicago, (2001) 2824. [6] M. Cornacchia et al. , Linac C oherent Light Source (LCLS) Design Study Report. Report SLAC - R-524 ( Stanford Linear A ccelerator C enter , Stanford, C alifornia, 1998); R. Brinkmann et al. , DESY Report DESY97-048 (Deutsches Elektronen Synchrotron, Hamburg, 1997). an energy f ilter. T he energy bandwidth wa s controlled by slight shift of the incident energy. A 4-quadrant s lit was employed to the extrac t s patially c oherent area of t he m onochromatic beam. T he true coincidence rate, C S , and th e a c c i d e n t a l o n e , C N , w e r e m e a s u r e d w i t h coincidence circuits and an electric delay of 4.79 μ s that corresponds to the revolution frequency of th e storage ring. T h e e n h a n c e m e n t R = C S / C N − 1 o f t h e coincidence rate is g iven by t he inverse of th e temporal m ode number M t , which m eans an average number of intensity fluctuation in a singl e pulse. We plotted t he m ode number M t as a function of the bandwidth ∆ E , as shown in . Assuming that the pulse envelop e and the temporal c o h e r e n c e p r o f i l e s a r e o f b o t h G a u s s i a n distributions, M t is given by [1+( s t / σ t ) 2 ] 1/ 2 from theory, where s t is the pulse width in FWHM and σ t is the coherence time given by (2 ln 2) h / ( π∆ E ) ( h is the Planck constant). T he data were fitted with one fitting parameter, that is, the pulse width s t . The width was determined to be 32. 7 ± 1.6 ps in FWHM . This value was compared to that measured with a streak camera, 32 ps [5]. This level of agreement was excellent . Intensity interferometry combined with various X-ray monochromators is capable of determining X- ray pulse widths in the timescale from ns down to fs. This is because the monochromators can cove r a wide range of bandwidth f rom 10 -4 to 10 eV , which corresponds to values of σ t between 10 ps and 0.1 fs. Im portantly , t he method can be easily extended to faster pulse regions because the optic s r equired are much simpler t han t hose used in t he present work. T he time resolution is unaffecte d by the timing jitter of the incident pulses and of th e trigger signal. This method provides a unique technique for characterizing ~ 100 - fs pulse profile s generated with t he forthcoming linac-based, coherent X-ray sources, in addition to much faste r X -ray pulses produced by proposed slicin g technique of chirped pulses or ultrafas t Bragg switches . 108 1 0 – 1 1 0 0 ∆ E ( m e V ) 1 0 0 1 0 1 M t
home
- JP
- EN