Prospects for IXS at an Energy Recovery LINAC light source
Abstract:
An Energy Recovery LINAC (ERL) based hard x-ray light source is being planned for construction at Cornell.
This 5GeV,100mA facility will generate x-ray beams of unprecedented average spectral brightness.
Although it is correct to think of ERL sources as optimized for high coherence and nanobeam science, they can
deliver ultra-high spectral flux because the small emittance and energy spread maximize efficient operation of very
long, short period undulators [1]. For example focused flux from a 20m ERL ID at 21.75KeV will be more than 100
that from a 3 ID upgrade considered for APS Sector 30 and at 9.1KeV (for CDW/CDDW optics) it would be almost 40
times that of a 5m NSLS-II U20 ID [2]. The ERL will enable pushing beyond the horizon of inelastic scattering studies
in biology, geology, chemistry and materials science. With a second injector, the Cornell ERL could simultaneously
feed an x-ray free electron oscillator (XFELO) [3]. The talk will introduce the ERL, provide an overview of facilities
with emphasis on IXS, and focus on new experiments that will become feasible. In particular we can envision: IXS
microscopy, selective measurements on single crystals within powder diffraction samples, advances in DAC based
high pressure science, and studies of new classes of biological systems.
Figure: Cornell ERL plan showing
major components and 14 undulator
beamlines, with 3 IDs up to 25m.
The experimental floor is indicated in
pink, while SC Linacs are in yellow.
References
[1] KD Finkelstein, et.al., J. Phys.Chem. of Solids 66, 2310 (2005).
[2] Estimate for 2-d focusing, 50m from source, based on specifications from each facility
[3] K-J Kim, Y Shvyd’ko, S Reiche, PRL 100, 244802 (2008).
Prospects for IXS at an Energy Recovery LINAC light source
K.D. Finkelstein, S. M. Gruner, G. Hoffstaetter, D. Bilderback
Cornell Laboratory for Accelerator-based Sciences & Education (CLASSE)
Outline of talk:
• Characteristics of an ERL pertinent to IXS
• Opportunity for long undulators: novel designs, beamline engineering
• Expected spectral characteristics in comparison to other sources
• Expanding the “phase space” for IXS measurements
• Niche applications for ERL-based IXS
• Summary & invitation
For ERL overview please visit poster N4: “The ERL: A Coherent, Hard X-ray Source”
Acknowledgements
To the entire ERL development team at:
Cornell Laboratory for Accelerator-based Sciences & Education (CLASSE)
Special thanks for continuing guidance on IXS for future ERL light source:
P. Abbamonte, A. Baron, B. Larson, M. McMahon, L. Pollack, K, Shen, Y. Shyvd’ko
Thank you to the organizers of this great meeting (IXS-2010) !
Characteristics of an ERL pertinent to IXS
ERLs are very flexible:
•
•
•
•
e- bunch properties determined by injector → short pulses, round beams
Table Error! No text of specified style in document.-1: ERL
operating mode target
- trajectory
Twissparameters.
parameters
flexible
&
adjustable
→
tailor
e
to optimize
th
An electron beam with a 4 set of characteristics for development
work photons
Simultaneous
with pattern
X-ray operation
in any of the for
above
modes,
a fast
flexible ise-envisioned.
bunch structure
& filling
→ opportunity
FEL
beamlines
kicker will pluck specific bunches out from the main stream at ≤ 10kHz. Bunch charge
no separate
constant
current,
very small
can be upinjection
to 1nC, inorbit
which→
case
the geometric
emittance
(h/v) istransverse
simulated to ID
be aperture
2600/37pm for an RMS bunch length of 100fs and a relative energy spread of 2.e-3.
Operating Modes
Energy (GeV)
Current (mA)
Bunch Charge (pC)
Repetition Rate (MHz)
Geom. Emittance (pm) h/v
RMS bunch length (fs)
Relative energy spread (1E-3)
A
High Flux
5
100
77
1300
30
2000
0.2
Standard run mode,
best for IXS
B
High Coherence
5
25
19
1300
8
2000
0.2
C - Short Bunch
North Arc | South Arc
5
25
19
1300
120/9
11/9
100
1000
2
These properties may be
compatible with XFELO
(at lower rep. rate)
FEL capabilities, compatible with routine ERL operation, being explored.
ERL opportunities –
long undulators,
simplify beamline engineering,
novel ID designs
1) Smaller dγ/γ & 2d emittance optimum for long undulators:
1000 period undulators should pay big dividends
2) Very narrow harmonics in energy and angle:
more ‘useful flux’ thru aperture & higher flux/Watt
3) New ID designs already under development:
Prototype “Delta undulator” built and tested (with 50-70MeV
electrons) at Accelerator Test Facility - Brookhaven National Laboratory.
1) Undulator harmonic energy width depends on:
length (N*λID), e- divergence (ε/β) , electron energy spread (δγ/γ)
for K~1, width is approximately*
δЕ/Е ~ √[ (1/nN)2 + (2 δγ/γ)2 + (γ2(ε/β)/(1 + K2/2))2 ]
3rd H. fractional energy width
0.05
0
200
400
600
800
1000
n is harmonic number.
ERL δγ/γ ~ 20% of storage rings
(dominates emittance term)
# periods
dE/E
APS
conditions
ERL hi flux
0.005
0.0005
ERL hi coh
* following Attwood
2) Benefit of narrow harmonics:
high ‘useful flux’ & modest power
(case of no focusing & combined harmonic power)
18mm period, 25m ERL “delta” ID;
3rd harmonic at 50m
"useful flux" @ 21.75KeV
1.2E+11
7.0E+10
20mm period, 4.5m Spring8 ID;
1st harmonic at 28m (Baron)
2.0E+10
0.2
0.4
0.6
0.8
1
1.2
Flux/Watt through aperture
600
2.4E+09
Power through aperture
photons/sec/mev/Watt
500
400
Watts
flux/meV
1.7E+11
300
200
100
0
0.2
0.4
for ERL-IXS
0.6
0.8
aperture area (mm2)
1
1.2
SPring8 BL35XU
2.1E+09
1.9E+09
1.6E+09
1.4E+09
1.1E+09
8.5E+08
6.0E+08
3.5E+08
1.0E+08
0.2
0.4
0.6
0.8
aperature area (mm2)
1
1.2
3) Novel IDs enabled by small, round e- beam size, no injection orbit
Delta ID: fixed gap, fully polarization tunable
Fields of Delta:
Helical: + 40%
Planar: +100%
SC - ID: offers higher fields → smaller period
&/or wider K-range (realization is challenging!)
Bmax [T] in helical mode
Period[mm]
Bmax [T] in planar mode
Delta
SC
Delta
SC
24
1.0117
1.097
1.431
2.194
22
0.9686
1.058
1.37
2.116
20
0.9162
1.009
1.296
2.018
18
0.8533
0.9497
1.207
1.899
16
0.7787
0.8679
1.101
1.736
Average spectral brightness
of ERL Delta ID (18mm period,
25m length) relative to existing
sources & NSLS-II U20, 3m ID.
Calculations based on Table 2
in New Journal of Physics 12,
035011 (2010) (on coherence
applications)
3rd harmonic flux-energy tuning range
( for 25m long “Delta” ID with 18mm period)
photons/sec/0.1%bw
1.E+16
3rd H flux (thru 1mm aperture@50m)
57Fe
1.E+15
1.E+14
1.E+13
12000
16000
20000
24000
Energy (eV)
28000
32000
36000
Covers full range of energies for high resolution IXS !
Where will ERLs shine ?
higher spectral flux:
boosts count rate from weak scattering systems
extends IXS to high Z materials
round beam emittance:
no drop in brightness with vertical polarization
no ‘blind spot’ for scattering near 90 degrees
large Q (horizontal scattering geometry)
studies at subatomic length scale
small ROUND source ideal for high throughput focusing:
extends practical range of high pressure studies,
IXS microscopy, select single grains within sample
brilliance, high peak current (short pulses) & small δE/E
are minimal requirements for XFELO, XFEL
short pulses might be used to:
reduce signal to noise of sample chamber (2ps<->0.6mm)?
pump-probe IXS ?
Coherent phonons ?
Spectral flux: ERL & potential future beamlines
units: 1014 p/s/0.1%bw
Resolution (meV)
Energy (KeV)
ERL “Delta” ID
λ=18mm, 25m,
1mm aperture @ 50m
SPring8 BL35XU
U20 – 4.5m
2
0.5 x 1.5mm @ 28m
ESRF ID28 @ 300mA.
3 Revolver IDs
2
0.6 x 1.6mm @ 27m
APS Sector 30
100mA.3 x U30 Ids
2
0.4 x 2mm @ 30m
NSLS-II baseline
500mA U20 5m hi-β
2
0.6 x 1mm @ 30m



<1 meV
(CDW optics)
9.1
6
Si(8 8 8)
15.82
1.5 (1.2)
Si(11 11 11)
21.75
0.9 (0.6)
Si(13 13 13)
25.7
118
(helical)
65
(3rd H)
35.4
(3rd H)
33.2
(5th H)
-
18
13
7.3
-
11.2
7.2
5.4
-
-
5.7
3.9
-
9.95
1.69
0.07
ERL Delta ID flux calculations assume helical mode below 12.4 KeV, planar above.
Other numbers are from AQR Baron except SPECTRA 8.0 calculations for NSLS-II based on
IXS@NSLS-II Feb.2008 workshop report by Yong Cai.
Aperture size & distance are characteristic for the beamline specified
Expand the “phase space” for IXS measurements
By delivering 6 - 12 times more spectral flux,
ERLs will make (R)IXS on high Z systems more practical.
from - ESRF ID 28
“Primer on IXS”
Ratio of total number of photons (Thomson) scattered to those lost through other
processes (predominantly photoelectric absorption) in sample of optimum thickness 1/μ
as a function of atomic number. (example for 1meV resolution studies).
Focused flux (photons/sec/μm2) at resolution specified
Resolution (meV)
Silicon (hkl)
Energy (KeV)
½
(CDDW optics)
9.1
6
(8 8 8)
15.82
3.7x1012
1.5
(11 11 11)
21.75
0.9
(13 13 13)
25.7
4.5x1011
1.8x1011
ERL 18mm Delta
(20m ID)
9.3x1011
APS Sector 30
100mA 3-U30 (7.2m)
-
-
3.5 x109
1.2 x109
NSLS-II baseline
500mA U20 hi-β
(5m)
2.5 x1010
5.3 x1010
2.33 x108
-
Spring8 BL35XU
U20 ID
(4.5m)
-
1 x1011
1.32 x1010
3.8 x109
ERL Delta flux calculations for helical mode below 12.4 KeV & planar above.
APS upgrade & Spring8 numbers from table provided by A.Q.R. Baron
NSLS-II based on Feb.2008 workshop & SPECTRA 8.0 calculation.
Numbers based on p/s/eV/mm2 50m from source, 2-d focusing at 100:1 (optics accept ½ mm by ½ mm)
ALL source sizes based on published emittance, beta, and photon energy
High Pressure DAC studies
Sample volume Т Asample is the region of uniform pressure
IXS signal
NIXS ~ I0(p/s/area) ρТ Asample (∂2σ/∂ΩЕ) ΔΩΔЕ e-μТ
[ρ scatterers/volume, absorption coefficient μ , ∂2σ/∂ΩЕ ~ |f(Q)|2 for single species]
D
IF Asample ~ D2
and T ~ D then
NIXS ~ D3 I0 |f(Q)|2
D is often [1] inversely proportional to pressure P so,
to obtain comparable signal
I0 ~ P3/ |f(Q)|2
The ERL delivering 100 times higher I0 (then today’s sources) will enable studies:
at up to 5 times higher pressure (for given |f(Q)| )
and/or
push frontiers for high pressure studies for low Z materials
For example the “holy grail” of low Z materials is hydrogen. The structure has been studied with x-rays to about 30GPa…
[1] A.L.Ruoff, H. Xia, Q. Xia, Rev Sci Instrum. 63, 4342 (1992)
The phase diagram illustrates how much further we must go to answer fundamental questions
about high pressure - temperature phases of hydrogen and some mixtures.
Potential for ERL-based pressure studies
Present limits in structural studies of hydrogen.
ERL could extend measurement in BOTH P & T!
Build in new science capabilities:
using TDS to guide IXS measurements
Duel energy resolution mono for switching incident beam between 1eV (for TDS) and 1meV (for IXS).
For 1eV downstream (partial blue) crystal of high resolution mono is removed and partial red inserted.
For 1 meV measurements arrangement is reversed.
Thermal diffuse scattering
(TDS) from silicon (a, b).
Model with (c, d) & without
(e, f) optical branches [2].
[2] R. Xu, T.C. Chiang, Z Kristallogr. 220, 1009 (2005).
Imaging nonequilibrium atomic vibrations with x-ray diffuse scattering
M. Trigo, Y. M. Sheu, J. Chen, V. H. Vishwanath, T. Graber, R. Henning & D. A. Reis
arXiv:1006.3990v2
A similar arrangement might be used to selectively measure oriented
single crystals within a powder
Possible optical arrangement for IXS DAC studies of crystallites located/oriented by Laue diffraction.
High heat load diamond mono (red) passes energy required for IXS. First (thin) diamond is misaligned relative
to second for beam transmitted to multilayer mono (blue outline). MML passes full harmonic width
(FWHM~60eV) for Laue-alignment of single crystal.
Refractive lens (green) focuses beam at DAC.
Diamond is realigned to direct beam to high resolution mono (HRM) & last HRM crystal (blue hatched)
must be accurately inserted to direct beam along MML beam path.
For Alfred:
Table C-2. Comparative Source Sizes and Divergences
Machine
Horiz. Horiz.
Vert.
Vert.
Size Diverge Size
Diverge
(μm)
(μrad)
(μm)
(μrad)
ESRF ID13 (4nm,
59
90
8.3
3
0.6% coupling)
APS-A (2.5 nm, 1%
275
11.3
8.8
2.9
coupling)
NSLS II (0.5 nm, 2%
28
19
2.6
3.2
coupling)
ERL, 25 m undul.
24.5
6.1
24.5
6.1
hi-flux mode; 30 pm
ERL, 1 m undulator
hi-coher mode; 8 pm
2
4
2
4
ERL experimental facility:
14 undulator beamlines
3 with IDs up to 25m long
IXS concept beamline –
25m undulator and 50m for hutches
space requirements based on SPring8 BL35XU
Potential location for XFEL-Oscillator
planned 25m ID
Summary and Invitation
Where the ERL will shine:
high-Z materials,
weak scattering (electronic excitations),
DAC and micro-sciences,
dynamics & phase transitions of liquids and glasses
The ERL will potentially support FEL extensions (XFELO ?)
WE need your interest and scientific & technical guidance.
Please attend workshops to be held at Cornell in June 2011…
Example:
Soft x-ray SASE FEL (optical length 50m)
kick subset of electron bunches from regular ERL beam to FEL line
compress to 100fsec
pass through special ID (several segments)
Yield:
x-rays at 1.86KeV (1st H)
100fsec pulses
bandwidth 0.12%
divergence 9.3urad
3.7e9 photons/pulse (peak brightness ~2e27)
Energy Recovery
Linac
Accelerating bunch
Returning bunch
A superconducting linac is required for
high energy recovery efficiency