Femtosecond electronic response of atoms to ultra-intense X-rays
(01 July 2010)
15 February 2010
10 May 2010
An era of exploring the interactions of high-intensity, hard X-rays with matter has begun with the start-up of a hard-X-ray free-electron laser, the Linac Coherent Light Source (LCLS). Understanding how electrons in matter respond to ultra-intense X-ray radiation is essential for all applications. Here we reveal the nature of the electronic response in a free atom to unprecedented high-intensity, short-wavelength, high-fluence radiation (respectively 10
18 W cm , 1.5–0.6 −2 nm, ~10 5 X-ray photons per Å ). At this fluence, the neon target inevitably changes during the course of a single femtosecond-duration X-ray pulse—by sequentially ejecting electrons—to produce fully-stripped neon through absorption of six photons. Rapid photoejection of inner-shell electrons produces ‘hollow’ atoms and an intensity-induced X-ray transparency. Such transparency, due to the presence of inner-shell vacancies, can be induced in all atomic, molecular and condensed matter systems at high intensity. Quantitative comparison with theory allows us to extract LCLS fluence and pulse duration. Our successful modelling of X-ray/atom interactions using a straightforward rate equation approach augurs favourably for extension to complex systems. 2
View full text
Figures at a glance
Figure 1: Diagram of the multiphoton absorption mechanisms in neon induced by ultra-intense X-ray pulses.
X-rays with energies below 870
eV ionize 2 s, p-shell valence electrons (V, red arrow). Higher energy X-rays give rise to photoemission from the 1 s shell (P, purple arrow), and in the consequent Auger decay the 1 s-shell vacancy is filled by a 2 s, p-shell electron and another 2 s, p electron is emitted (A, black arrow). These V, P and A processes are shown in more detail in the inset; they all increase the charge of the residual ion by one. Main panel, three representative schemes of multiphoton absorption stripping the neon atom. The horizontal direction indicates the time for which atoms are exposed to the high-intensity X-ray radiation field, and vertical steps indicate an increase in ionic charge due to an ionization step, V, P or A. Horizontal steps are approximately to scale with a flux density of 150 X-ray photons per Å 2 per fs, and indicate the mean time between photoionization events or Auger decay.
Figure 2: Neon charge-state yields for X-ray energies below, above and far above the 1 s-shell binding energy, 870 eV.
Pulse energies are measured in the gas detector upstream of the target.
a, Experimental charge-state distribution for 2.4-mJ pulses at 800 eV (top), 1,050 eV (middle) and 2,000 eV (bottom). b, Comparison of experimental charge-state yields, corrected for detection efficiency, with simulations, assuming a Gaussian-shaped pulse, as described in ref. 12. The X-ray pulse durations for the simulation are assumed to equal the electron bunch duration, shown in the figure. Fluence-dependent processes, such as valence stripping at 800 eV, are insensitive to the pulse duration.
Figure 3: Intensity-induced X-ray transparency.
a, Ratio of charge-state yield for 230-fs to 80-fs electron bunch durations for 2,000-eV, 2.0-mJ X-ray pulses (diamonds). Error bars of ±10% are estimated for the observed charge-state yields. Simulations for 80-fs and 20-fs X-ray pulse durations for the short pulse are overlaid in green and red, respectively. The comparison suggests that the pulse duration of the X-rays generated by the 80-fs electron bunch is less than 80 fs. b, Average lifetimes of the single- and double-core-hole states in neon as a function of charge state, from ref. 33. The lifetimes of the single- and double-core holes increase with charge state. These increased lifetimes give rise to a decreased absorption cross-section and hence an intensity-dependent X-ray transparency. The double-core-hole lifetime tracks the observed charge-state ratios. c, A scheme leading to high charge states, for a flux density of 2,000 X-ray photons per Å 2 per fs, corresponding to that on target for a 2.0-mJ, 2,000-eV, 20-fs X-ray pulse.
Figure 4: Electron spectra for inner-shell and valence photoelectrons and Auger electrons created by X-ray pulses at a photon energy of 1,050 eV.
a, 1 s photoelectrons directed along the X-ray polarization axis ( = 0°) with a photoelectron spectrometer retardation voltage of θ V R = 0 V. b, Valence and Auger electrons at = 0°, with θ V R = 790 V. c, Valence and Auger electrons at = 90°, θ V R = 790 V. Auger electrons from double-core-hole states, the signature of hollow atom formation, are more cleanly measured at = 90° (as in θ c), as background from valence ionization is suppressed. The strongest Auger peaks in the double-core-hole and single-core-hole regions originate from the initial PPA and PA ionization sequences, and are used to determine the double-core-hole to single-core-hole formation ratio.
Figure 5: The ratio of double- to single-core-hole formation as a function of X-ray pulse energy.
Red filled circles show experimental data taken with X-rays lasing from an electron bunch of 250
pC charge and 80 fs duration. A gas attenuator was used to reduce the pulse energies. Error bars are statistical (±1 ) for the Auger yields and measured shot-by-shot X-ray pulse energies. Curves labelled 20 σ fs (green) and 40 fs (blue) are simulations of the double-to-single core hole formation ratio for two X-ray pulse durations. The comparison suggests that the 80-fs electron bunch generates an X-ray pulse of shorter duration, ~20–40 fs.