Spectroscopy of Highly Charged Ions with Free Electron Lasers. Sascha Epp MPI-K Heidelberg & ASG within CFEL PSAS 2008

Transcription

1 Spectroscopy of Highly Charged Ions with Free Electron Lasers Sascha Epp MPI-K Heidelberg & ASG within CFEL PSAS 2008

2 People involved in this work MPI-K S. W. E. (ASG) J. R. Crespo L. U. G. Brenner V. Mäckel M. Simon T. Baumann P. Mokler R. Ginzel H. Tawara J. Ullrich DESY R. Treusch M. Kuhlmann N. Guerassimova M.V. Yurkov E. Schneidmiller J. Feldhaus J. R. Schneider University Hamburg M. Wellhöfer M. Martins W. Wurth Support by Max-Planck Gesellschaft: Advanced Study Group (ASG) at the Center for Free Electron Laser Physics (CFEL)

3 α. Motivation - Laser spectroscopy - S.W. Epp et al., Phys. Rev. Lett. 98, (2007). H 1S-2S 1. Bi 82+ I. Klaft et al. 5.1 ev (160 ppm) 2. S 15+ H. Sträter et al. 2.9 ev (200 ppm) 3. F 7+ E. Myers et al ev (1 ppm) Hänsch, T. First demonstration of resonant laser spectroscopy (l. s.) of a bound transition in the soft x-ray regime. Extending l. s. to a new class of targets ground state transitions in Highly Charged Ions (HCI) where most transitions lay in the (soft) x-ray regime.

4 α. Motivation -Astrophysics - HCIs constitute a dominant fraction of the visible matter in the universe! Sun Active Galactic Nuclei Comets

5 α. Motivation - Atomic structure theory - Relativistic corrections (Dirac) Ζ 4 H-like atom E n Z 2 QED corrections Ζ 4 Electron correlations Ζ 1 Corrections are boosted in HCIs compared to neutral systems. scaling laws Electron correlations are suppressed. HCIs are an ideal testing ground for theory.

6 H-like α. Motivation - Atomic structure theory - Li-like H-like Li-like Energy Lamb shift QED QED QED QED Lamb shift QED contributions Li: 0.002% of 1.85 ev Fe 23+ : 1% of 48.6 ev U 89+ : 15% of 280 ev

7 β. Experimental setup FLASH (Free electron LASer in Hamburg) FLASH-EBIT (electron beam ion trap) Accelerator hall H = 2.4 m L = 2.5 m M = 1.5 t Undulator tunnel 240 m 240 m Experimental hall

8 β. Experimental setup - FEL principle - Radiation field interacts with electron bunch and structures it Electron beam Radiation

9 β. Experimental setup - FEL principle - Coherent superposition enhances intensity by a factor proportional to the number of charges in the bunch N e Dipole Wiggler Undulator Undulator Brilliance increases by many orders of magnitude

10 β. Experimental setup - FLASH-EBIT -

11 β. Experimental setup - FLASH-EBIT - First skillful task: Overlap of ion cloud and photon beam Mirror

12 β. Experimental setup - FLASH-EBIT -

13 γ.experiment & Results - Laser spectroscopy - FLASH: Ava. spectral distribution FLASH: Repetition structure Use monochromator E 0 /F.W.H.M 2,000 (>20,000 possible) 5 Hz repetition rate =>150 single pulses/s Average power : up to 10 mw (laser pointer 4 mw) at time of experiment S P Fe 23+ τ 0.6 ns Laser 48.6 ev Resonant fluorescence 2 2 (1s 2s) 2 S1/2 2 2 (1s 2p) 2 P3/2 2 2 (1s 2p) 2 P1/2

14 γ.experiment & Results - Results 2007 (1 st run) - Arrival time (μs) Fe S 1/2-2 2 P 1/2 Integral counts on MCP(E, t) right scale FIT (Gaussian): 60 Ex 0c = (11)(150) background ev! 370 true counts in 5σ with 50 baseline subtraction 40 =>Use FLASH time structure ϰ 2 /dof! = 1.23 FWHM=0.0353(24) ev 30 E 0 /FWHM = 1350 (750 ppm) => don t count all the time Photon energy (ev) ±0.015 ev Counts 33 minutes acquisition time different colour => diff. amount of counts We have to separate a 0.2 cts/s true 70 fluorescence signal from a cts/s no FLASH-light => no fluorescence E 0 = ( ±0.0011) ev => use gate δ = = ppm Systematic shift due preliminary calibration of monochromator unit! S.W. Epp et al., Phys. Rev. Lett. 98, (2007).

15 γ. Experiment & Results - Calibration parenthesis - How much is e.g ev? equivalent to THz defined by the second: The second is the duration of periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the cesium 133 atom. no frequency chain goes into the deep VUV, soft x-ray range frequency comb recently extended to 60 nm (20.7 ev) A. Ozawa et. al. PRL 100 (2008) calibration is a challenging task! photo-ionization resonances of noble gases desirable: highly accurate theoretical data of light hydrogen- and helium-like ions. (B, C, N, O, F) => wavelength standard in the soft x-ray

16 γ.experiment & Results - Results 2008 (2 nd run) Fe S 1/2-2 2 P 1/2 counts Data: Fescan106Mono_B Equation: y0+a*exp(-4*ln(2)*((x-xc)/fwhm)^2) Weighting: y Statistical Chi^2/DoF = R^2 = y ±0.74 xc Li- like Fe 23+ 2s-2p 1/ ev by 1 mev (256 x binned) 3600 s run ± (8 ppm) FWHM ± A ±4,46 uncalibrated raw data 0 48,52 48,53 48,54 48,55 48,56 48,57 48,58 48,59 48,60 Energy (ev)

17 γ. Experiment & Results -Neon Calibration - Photoabsorption signal (arb. u.) xc ± w 0.016±0.002 xc ± w 0.012±0.002 xc ± w 0.016±0.001 Wilhelmi et al., Journal of Electron Spectroscopy and Related Phenomena (1999) mev offset of the Mono. energy scale with an estimated accuracy of ±0.4 mev -200 Signal MCP Signal YAG Average Photon energy (ev)

18 γ.experiment & Results - Results 2008 (2 nd run) - counts Equation: y0+a*exp(-4*ln(2)*((x-xc)/fwhm)^2) Weighting: y Statistical Chi^2/DoF = R^2 = y ± xc ± (8 ppm) FWHM ± A ±2.571 Fe S 1/2-2 2 P 3/2 Li- like Fe 23+ 2s-2p 64.5 ev (19.2 nm) 3600 s run S Fe 23+ Laser 48.6 ev P Resonant fluorescence 2 2 (1s 2s) 2 S1/2 2 2 (1s 2p) 2 P3/2 2 2 (1s 2p) 2 P1/2 10 uncalibrated raw data Photon energy (ev)

19 γ.experiment & Results - Results 2008 (2 nd run) - 30 Cu S 1/2-2 2 P 1/2 25 Li- like Cu 26+ 2s-2p 1/2 C S P 0,1,2 Cu 26+ 2s-2p 1/2 Chi^2/DoF = R^2 = counts Chi^2/DoF = R^2 = y ± xc ± w ± A ± y ± xc 55.1xxx (± ) ±0.004 w ± A ± Energy (ev) uncalibrated raw data

20 γ. Experiment & Results 10 0 Contributions 10-1 Screening SE O(α 2 ) SE + VP O(α 2 ) - 3 photon inter-electron 3 photon inter-electron E (ev) Source of main error Nuclear size Recoil Screening SE + VP O(α 3 ) 3 & more photons inter-elec. Theory Yerokhin et. al. '08 Experiment Other Lestinsky et. al. 2s-2p 3/2 This work '07 (laser spec.) This work '08 (laser spec.) 8 ppm in 3600 s Nuclear charge Z

21 δ. Summary demonstration of resonant laser spectroscopy of HCI in an EBIT with soft x-rays at photon energies as high as 65 ev using FELs - advantage: selective, resonant huge cross sections nearly routinely 8 ppm statistical precision in 1 h of D.A.Q. -> unprecedented accuracy is expectable working on calibration to convert precision in absolute accuracy - photo-absorption in noble gases - comparison with transitions in He- & H-like ions (Cu 26+ ) (9) ev (Reader, J. et al. ( 94)) 2008 Literature Fe (11)(150) ev 2007

22 δ. Outlook FLASH laser synchrotron light sources LCLS (Lin. Collider Light SLAC Stanford European X-FEL in Hamburg FLASH ev ev ev BESSY ALS intra-shell (n=n ) transitions LCLS (SXR) ev ( ev) ev ALS inter-shell (n<n ) t. photo-ionization X-FEL ev ev ev ev APS ESRF PETRA III inter-shell (n<n ) t. photo-ionization

23 δ. Outlook FLASH - FLASH still misses 2-3 orders of flux compared to design values -> more STATISTICS FLASH one order of mag. more resolution power possible 2k->20k Monochromator with up to 150k exist Seeding of FLASH will enhance #photons/ev B.W. systematic studies of Li- or Be-like HCIs over a wide Z- range, including isotopes => a consistent body of data to resolve nuclear and QED effects

24 δ. Outlook synchrotron light sources - FLASH still misses 2-3 orders of flux compared to design values -> more STATISTICS FLASH Problem: Time structure! Synchrotrons are approx. c.w.

25 δ. Outlook synchrotron light sources - What we do: Velocity filter LCLS photon beam Photoionization meassurements Last week: N 3+ -> N 4+ BESSY, Berlin Position-sensitive ion detector (TOF) EBIT Ion-photon interaction volume Fluorescence and spectral diagnostics Ion deflector HCI beam Ion optics Beam imaging & diagnostics Electron collector Trap electrodes Electron gun

26 δ. Outlook synchrotron light sources - Intensity (counts) Energy (ev) EBIT<-> PHOBIS Intensity (counts) 700 Model: Fano Weighting: Statistical Chi^2/DoF = R^2 = The use of narrow photoionization resonances of Ne (He) etc. can provide calculable standards in the region up to 80 ev q ± 0.04 Γres ± Eres ± y0 161 ± 4 A 107 ± 5 L Fano simulation residue Photon energy (ev)

27 δ. Outlook LCLS from first experiments: ev photon energy Photo-ionization of HCI (energies & cross sections) e.g. Fe-Fe 23+ => l.s. of the sun in the EBIT He-like Systems at the edge of performance: H-like 1s-2s lifetime measurements - elaborated scheme with utilizing fs time structure (difficult) - simpler scheme for the few ns range (immediately doable)

28 δ. Outlook LCLS from counts (arb. units) 30k 25k 20k 15k 10k 5k fluorescence time coincidence counts (arb. units) fluorescence time coincidence Data: scan003tdc001_b Model: Gauss mit FWHM Equation: y0+a*exp(-4*ln(2)*(x-xc)^2/fwhm^2) Weighting: y Statistical Chi^2/DoF = R^2 = time mark y0(ns) ± xc ± FWHM ± A ± lifetime measurements targeting the few ns range advantage: high photon fluxes only few measurements, not better than 10% acc time mark (ns)

29 δ. Outlook X-FEL from Possibilities increase even more for transitions mentioned and not mentioned in this talk H-like 1s-nP etc. full accessible for lot of HCIs Targeting 1s Lamb-shift by laser spectroscopy Wavelength standard by HCIs?

30 δ. Outlook LCLS from