LUSI XPCS Status

1
LUSI XPCS Status
Team Leader Brian Stephenson (Materials Science
Div., Argonne) Co-Leaders Karl Ludwig (Dept. of
Physics, Boston Univ.), Gerhard Gruebel
(DESY) Sean Brennan (SSRL) Steven Dierker
(Brookhaven) Eric Dufresne (Advanced Photon
Source, Argonne) Paul Fuoss (Materials Science
Div., Argonne) Randall Headrick (Dept. of
Physics, Univ. of Vermont) Hyunjung Kim (Dept. of
Physics, Sogang Univ.) Laurence Lurio (Dept. of
Physics, Northern Illinois Univ.) Simon Mochrie
(Dept. of Physics, Yale Univ.) Larry Sorensen
(Dept. of Physics, Univ. of Washington) Mark
Sutton (Dept. of Physics, McGill Univ.)
LCLS SAC Meeting June 7-8, 2006
2
Scientific Impact of X-ray Photon Correlation
Spectroscopy at LCLS

• New Frontiers
• Ultrafast
• Ultrasmall
• Time domain complementary to energy domain
• Both equilibrium and non-equilibrium dynamics

3
Unique Capabilities of LCLS for XPCS Studies

• Higher average coherent flux will move the
  frontier
• smaller length scales
• greater variety of systems
• Much higher peak coherent flux will open a new
  frontier
• picosecond to nanosecond time range
• complementary to inelastic scattering

4
Wide Scientific Impact of XPCS at LCLS

• Simple Liquids Transition from the hydrodynamic
  to the kinetic regime.
• Complex Liquids Effect of the local structure
  on the collective dynamics.
• Polymers Entanglement and reptative dynamics.
• Proteins Fluctuations between conformations,
  e.g folded and unfolded.
• Glasses Vibrational and relaxational modes
  approaching the glass transition.
• Dynamic Critical Phenomena Order fluctuations
  in alloys, liquid crystals, etc.
• Charge Density Waves Direct observation of
  sliding dynamics.
• Quasicrystals Nature of phason and phonon
  dynamics.
• Surfaces Dynamics of adatoms, islands, and
  steps during growth and etching.
• Defects in Crystals Diffusion, dislocation
  glide, domain dynamics.
• Soft Phonons Order-disorder vs. displacive
  nature in ferroelectrics.
• Correlated Electron Systems Novel collective
  modes in superconductors.
• Magnetic Films Observation of magnetic
  relaxation times.
• Lubrication Correlations between ordering and
  dynamics.

5
XPCS using Sequential Mode

• Milliseconds to seconds time resolution
• Uses high average brilliance

sample
transversely coherent X-ray beam
monochromator
movie of speckle recorded by CCD
1
6
transversely coherent X-ray pulse from FEL
7
XPCS of Non-Equilibrium Dynamics using Pumped
Mode

• Femtoseconds to seconds time resolution
• Uses high peak brilliance

before
sample
transversely coherent X-ray beam
?t after pump
monochromator
Pump sample e.g. with laser, electric, magnetic
pulse
Correlate a speckle pattern from before pump to
one at some ?t after pump
8
Split Pulse - Sequential Mode Crossed Beams
Femtoseconds to nanoseconds time resolution Uses
high peak brilliance
sample
transversely coherent X-ray pulse from FEL
variable delay
Crossed beams at sample allows recording of
separate speckle patterns from prompt and delayed
pulses (SAXS from 2-D samples)
1
9
Design of Experiments
Driven by analysis of sample heating by beam For
these studies of dynamics, we must avoid changing
the behavior of the sample by the beam (e.g. lt 1K
heating)
10
Sample Heating and Signal Level
Is there enough signal from a single pulse? Is
sample heating by x-ray beam a problem?
Maximum available photons per pulse
Minimum required photons per pulse to give
sufficient signal
Maximum tolerable photons per pulse due to
temperature rise
See analysis in LCLS The First Experiments
11
Heating and XPCS Signal from Single Pulse
Shaded areas show feasibility regions e.g. for
liquid or glass (green) or nanoscale cluster
(yellow)
See analysis in LCLS The First Experiments
12
Optimum pixel size 1 speckles
Detector Specifications
Pixel Size, Noise Level, Number of Pixels,
Efficiency Speckle negative binomial
distrib. Mean counts per pixel Inverse contrast
M Probability of k counts
Required signal/noise determine P2 to a few
need N2 Ntot k2 gt 1000
Low count rate limit
Required Ntot (number of pixels at same Q) 106
to 108
13
Current Detector Questions
1) In order to get large number of pixels, need
to understand trade-offs between number of
pixels, pixel size, noise level, efficiency,
cost Can an inexpensive commercial technology be
adapted? 2) For XPCS, pixels do not have to be
contiguous. Using a mask to separate pixels
could be a flexible way to produce small pixels,
and reduce noise due to charge sharing between
pixels
14
Beam Size at Sample

• Larger gives less heating per total signal, but
  size limited by ability to resolve speckle
  pattern in reasonable sample-to-detector distance
• Beam size pixel size speckle size d
  (?L)1/2
• For L 5 m, get
• d 20 microns, 8 keV d 12 microns, 24 keV
• Unfocused beam size at 8 keV is 400 microns
• Can use large coherent beam to
• - split beam spatially to produce time delay
• doing heterodyne detection using reference beam
• feed another experiment

15
Conceptual Design Mono and Splitter

• Si (220) or C (111) energy resolution typ., 6-24
  keV
• Pulse splitter - 3 concepts
• Partially-transmissive reflection e.g. Laue
• Split energy spectrum
• Split spatially (should be 100 m upstream to
  combine at minimum angle)
• For times longer than 1 ns, should consider two
  pulses in linac
• Mono upstream of splitter would remove heat load
  and avoid any effect of first pulse on second

16
Conceptual Design Beamline Layout

• Hutch in far hall
• 10 m long by 10 m wide hutch, with slits
  upstream for SAXS region, 15 m long would be
  more flexible
• Need very low background (mirror system in front
  end will solve)
• Concerned about stability of upstream optics
  (need 0.5 microradian)
• Either no focusing or moderate (up to 11),
  compound refractive lenses in upstream tunnel
• Pumped mode experiments will require synchronized
  lasers

17
Conceptual Design Beamline Layout
Far exp. hall
Hutch
10 m
Defining apertures
Detectors
Pulse Splitter
Focusing Optics
Horiz. offset monochromator
Transmitted Beam
15-20 m
Sample
100 m
18
Large Offset Monochromator
XPCS requires monochromator Mono offset can be
used to separate beams, eliminate 'flipper'
mirrors Transparent first crystal could allow
simultaneous operation of other station(s)
Goniometer and Sample Chambers
Plan 3 different chambers for different T
regions Flight paths and detector supports
require thought
19
Summary of RD Needs, Sub-Teams

• Detector and Algorithm (Lurio, Mochrie)
• Split/Delay (Gruebel, Stephenson)
• Beam Heating of Sample (Stephenson, Ludwig)
• Large Offset Mono (Stephenson, Gruebel)
• Goniometer and Sample Chamber (Ludwig, Sutton)

20
Multilayer Laue Lenses A Path Towards
One-Nanometer Focusing of Hard X-rays
Multilayer Laue Lens
Deposition of thick, graded multilayer at APS
sectioning and microscopy at MSD/EMC/CNM.
WSi2/Si, 728 layers 12.4 mm thick
Dr10 nm
Dr58 nm
Electron microscopy shows accuracy of layer
spacings

• Theory
• An ideal Multilayer Laue Lens should focus X-rays
  to 1 nm with high efficiency.

• Experiments
• We have fabricated partial MLLs and measured
  their performance. The results support the
  predictions of theory.

H. C. Kang, G. B. Stephenson, J. Maser, C. Liu,
R. Conley, S. Vogt, A. T. Macrander (ANL)
21
Sub-20 nm Hard X-ray Focus
Section depth 13.05 mm, Drmin5nm, f2.6 mm
_at_APS 12BM
FWHM 19.3 nm E 19.5 keV h 33