Download 9.1.4 Other common usage diagnostics

9.1 Commissioning and common usage diagnostics
9.1.1 Requirements
The diagnostic requirements are derived from the User experimental requirements,
based primarily in the LCLS Initial Experiments. (including FEL commissioning, and
facility requirements (see section 9.3). Tables of requirements for each of the initial
experiments have been compiled, those required for FEL commissioning, and those
requested by multiple Users, are separated and described here (atomic physics, plasma
physics, FEL physics, nanoscale dynamics in condensed matter, femtosecond chemistry,
structural biology). Other experiment-specific diagnostics are dealt with in the sections
describing each of the first experiments. Based on these tables, the diagnostics are
considered as being required for 827 eV (1.5 nm) first harmonic operation and 8.28 keV
first harmonic operation (0.15 nm), and allowing for measurement of both the coherent
and spontaneous parts of the spectrum.
Present day x-ray detection techniques and time resolutions are: inorganic scintillators
(10 ns) and plastic scintillators (1 ns), semiconductor pn junctions (100 ps),
semiconductor photoconductors (6 ps) and photocathodes (20 ps). Thus present day
techniques have a best time resolution that is about the inter-pulse time, but not the pusle
duration of ~200 fs, let alone the internal ~1 fs spikes. That is, only parameters totalled
over an individual pulse can be obtained as long as the data is read out before the next
pulse arrives. The technology nearest to that required is the semiconductor
photocathodes. One solution to obtaining sub 100 fs resolution may be electronic or
optical preprocessing of all or part of the beam, but then phase information will be lost.
9.1.2 Coherent radiation properties
9.1.2.1 Spectrum
9.1.2.1.1 low energy: grating spectrometers
Roman Tatchyn is providing this section
9.1.2.1.2 high energy: crystal spectrometers
Roman Tatchyn is providing this section
9.1.2.2 Total energy (including profile)
9.1.2.2.1 Total energy, accurate, intrusive: calorimetery
John Arthur is providing this section
9.1.2.2.2 Total energy, every pulse, non intrusive: ionization chamber
The total energy can be measured using an ionization chambers. The proportionality
between ionization current and photon number breaks down at high incident photon flux,
because space charge is important. The limiting factor is the very low ion mobility. The
ions experience resonant charge exchange on the atoms, the cross-section is very large
and, accordingly, the mobility is very low. So, in order to prevent a build-up of the
positively charged column, one has to apply very high voltages. This, in turn, is limited
by the Paschen breakdown constraint. Helium is better than the heavier gases because of
a favorable (1/M) dependence of the mobility on the ion mass, a smaller charge exchange
cross-section (which also leads to an increase of the mobility) and a higher breakdown
limit.
9.1.2.2.3 Profile, intrusive and partially intrusive: x-ray CCD camera
The LCLS X-FEL needs x-ray imaging systems for measuring the spatial distribution
and divergence of the raw beam, for alignment and focusing of optical elements, and for
pulse-to-pulse monitoring of the beam shape, centroid, and intensity. Unfortunately a
single instrument cannot currently meet these requirements.
The most useful instrument, in the short term, will be a high-resolution, CCD-like,
camera for measuring spatial distributions and for alignment and focusing of optical
elements. Traditional instruments have used phosphorus screens to convert x-rays to
visible light that can be recorded by a CCD. Even with a microscope objective to
magnify the screen, the spatial resolution is limited by the spatial resolution of the
phosphorus that is typically in the range of 10 to 50 microns. Such resolutions are of
marginal utility to the LCLS which has a beam diameter at 8 keV of 100 microns.
Recently workers at the ESRF synchrotron facility have discovered a crystal that when
used as a scintillator for x-rays achieves 0.8 micron resolution. The scintillator is a 5
micron thick Ce doped YAG crystal on a 100 micron YAG substrate.
This camera will not work as a pulse-to-pulse beam monitor device at LCLS because
of its slow readout speed, and the susceptibility of the YAG crystal to damage in the
intense X-FEL beam. Nevertheless the YAG crystal technology could be utilized for this
purpose with slight modification as shown in figure 9.1.2.1.
FEL Beam
Objective
Be
Reflected beam
YAG
CCD
Figure 9.1.2.1 Concept for a pulse-to-pulse beam monitoring camera. A small amount
of the x-ray FEL beam is reflected by the Be foil onto a YAG scintillator whose emissions
are recorded by a fast framing CCD camera.
In this concept a thin foil of a low Z material such as Be acts as a beam splitter to
partially reflect a portion of the beam onto the YAG crystal. The YAG crystal imaging
system is the same as before except a fast framing CCD is substituted for the large format
CCD used above. The major foreseeable problem with this concept is the background xray radiation impinging on the crystal due to Compton scattering of the FEL beam by the
Be foil, and any fluorescence from an oxide layer on the foil surface. The existence of
fast framing CCDs with suitable format and frame rate for the LCLS is not an issue, as
suitable fast framing CCDs exist today.
An alternate proposal is to utilize Compton scattering of the beam off of a low Z gas
or solid, utilizing the crystal/fast framing camera.
Gas s cattering cell
FEL Beam
Slits
Compton
s cattered
x rays
YAG
Ob jective
CCD
Reflected b eam
Figu re 1 APulse-to-pulse
itor ing ofto beam
pro file
utilizwidth
ing Comp
tonCompton
Figure 9.1.2.2
schematic ofmon
a system
measure
beam
using
scattering from a gas cell
9.1.2.3 Centroid location
9.1.2.3.1 Ionization chambers
The centroid location can be measured by multiple, suitably instrumented, ionization
chambers, (also see the gas cell attenuator (9.2.2.1.3). Position detection is accomplished
using segmented electrodes (K. Sato, Proceedings of the SPIE, detectors for
cystallography and diffraction studies at synchrotron sources, Vol 3774 p 114 (1999)).
Using 4 cm long electrodes, with a sawtooth base of 3 mm and height 10 mm, and low
noise current read-out, the horizontal and vertical centroid can be measured to < 10 m.
9.1.2.3.2 crossed wire arrays
Richard Bionta is providing this section
9.1.2.3.3 mirrors
This measurement is performed by the ‘less intrusive profile diagnostic’, based on a
CCD camera, see section 9.1.2.2.2.
9.1.2.4 Divergence
This measurement is performed by multiple versions (at different locations) of the
less intrusive profile diagnostic, based on a CCD camera, see section 9.1.2.2.2.
9.1.2.5 Temporal distribution
One proposal to measure the overall pulse length is found in [B. Adams, Nucl.
Instrum Meth., A459 339 (2001)]. An intense fs laser is used to depopulate the valence
band of a semiconductor. Then incident x-rays with a particular nergy would be
absorbed. Recording of a secondary process such as Auger electron emission, x-ray
fluorescence, or resonant Raman scattering may allow the pulse duration to be measured.
9.1.2.6 Temporal Coherence:
9.1.2.6.1. Asymmetric Michelson interferometer
Time-domain autocorrelation can be used to obtain information on the power spectral
density of a temporal signal. Specifically, it can provide information on the temporal
coherence length of a quasi-coherent source [1]. In the case of the LCLS, this is a
critically important parameter directly related to the FEL-induced microbunch structure
in the electron beam [2], and provisions for characterizing its statistics are presently
under study by the LCLS X-Ray Optics R&D group [3]. With regard to interferometer
design, a basic requirement is to minimize distortions in the temporal structure of the
pulse so that valid autocorrelation spectra can be generated. A second requirement,
arising from the extreme power density of the LCLS pulses is that the response of an
optical element interacting with an LCLS pulse should remain sufficiently uninfluenced
by the energy being absorbed during the interaction.
A schematic of the proposed instrument is shown in Fig. 9.1.2.7.1. Although a
transmission grating splitter is explicitly shown, techniques based on
reflecting/transmitting foils [4] could also be considered. The split beams are reflected
off mirrors M1 and M2 and recombine at the detector plane. With the mirrors parallel the
path length difference between the two interferometer arms is 0. To induce a path length
difference M1 is rotated counterclockwise through an angle . At the same time, the
detector is rotated through the same angle and translated back a distance r2.
Figure 9.1.2.7.1 Schematic layout of an asymmetric Michelson autocorrelator based
on a grating splitter.
As  is tuned, the path difference r between the two arms is given by
sin( 1 )sec( 1  2 )  tan(1 ) 
r  (r0  r2 )  r1  2r0  

 tan( 1 )  tan(1  2  )

and the stroke distance r2 of the scanning/rotating detector by
(1)
tan(1 )  tan(1  2 ) 
r2  r0  

tan(1 )  tan(1  2 ) 
(2)
To illustrate parameter dependence, expand r about  = 0. This yields:
r  (2r0 1 )   (2r0 12 )  2  ...
(3)
 

 2
 2 2  41
r2  r0  
1  1 
 ...


(




)
3
3
(




)
 1

1
(4)
and
To illustrate a typical parameter range, let r0 = 0.5m. In Fig. 9.1.2.7.2 the variables r2
and r are plotted as functions of  for two values of 1.
r2 [m]
r [] (x100)
2
1.5
r2 (Theta 1 = 0.005r)
r2 (Theta1 = 0.01r)
Delta r (Theta 1 = 0.01r)
Delta r (Theta 1 = 0.005r)
1
0.5
0
0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045
 [rad]
Figure 9.1.2.7.2 Tuning curves for an asymmetric Michelson autocorrelator for two
1. r0=0.5m. The ordinate values for each curve scale
linearly with r0.
For a given starting value of 1 (viz., in the symmetric limit), the reflectivities of M1
and M2 are equal. As the device is tuned, the reflectivity of M1 will increase (assuming
it is a specular reflector), inducing an asymmetric scaling in the relative amplitudes of the
interferometer beams. In this regard, the maximal asymmetry, as well as the maximum
power loading condition, for specular interferometer facets will be determined by the
mirror material and 1. To illustrate the range of Z-dependent reflectivities accessible
with specular mirror materials curves spanning the 1 keV - 8.5 keV range of the LCLS
fundamental are plotted vs. 1 for gold and beryllium in Fig. 9.1.2.7.3. Corresponding to
these curves, the energy loading [eV/atom] induced in the mirror material - assuming
irradiation with the full non-attenuated LCLS beam - is plotted in Fig. 9.1.2.7.4. It is
evident that in order to develop a value of r of the same order of length as the LCLS
pulse without excessive values of r0 or r2 the preferred value of 1 will lie in the > 0.005
rad range. In order to operate in this range, it consequently follows that substantial
attenuation of the LCLS beam will be desirable prior to reflection in the interferometer.
In the present design this can be accomplished in part by the grating splitter, whose
diffraction efficiency can be controlled by adjusting the parameters (primarily the
thickness) of the transmission grating. Without upstream attenuation the grating itself is
likely to be damaged; however, a new grating, or grating area, could be inserted between
shots, which would also allow for spectral tuning. The basic requirements on the grating
parameters is that the dispersion be small enough to preserve the temporal structure of the
LCLS pulses and that the spectral bandwidth of the diffracted orders be substantially
larger than the bandwidth of the LCLS radiation.
1
Absolute reflectivity
Au 1 keV
0.8
Au 8.6 keV
0.6
Be 1 keV
0.4
Be 8.6 keV
0.2
10 -5
0.0001
0.001
0.01
0.1
Grazing Incidence Angle [rad]
Figure 9.1.2.7.3 Absolute reflectivities of Au and Be vs. grazing incidence angle at
photon energies of 1 and 8.6 keV.
100
 A[eV/atom]
10
1
Au 1 keV
Au 8.6 keV
Be 1keV
0.1
Be 8.6 keV
0.01
0.001
0.0001
10 -5 -5
10
0.0001
0.001
Grazing Incidence Angle
0.01
0.1
Figure 9.1.2.7.4 Energy loading of Au and Be vs. grazing incidence angle at photon
energies of 1 and 8.6 keV. Irradiation with the unattenuated LCLS beam is assumed.
Issues include:
a) Path length and resolution. The practically attainable path length difference of
the instrument is of the order of 20 m, for interferometer lengths of 2 m or
less. However, since the resolution is given by dn/n ~ (l/r), this still
corresponds to a resolution of the order of 5x10-6 at 8.5 keV. For larger
values of 1 or substantially longer arm path lengths, correspondingly longer
path length differences (and resolutions) could be generated.
b) Recording. As the path length difference is tuned, the detector need not record
the detailed fringe pattern associated with a given pulse. A mask consisting of
apertures smaller than ~ l/(8(1-)), and with a varying period equal to the
varying fringe pattern period, followed by an intensity detector would be
adequate, provided the contrast ratio of the interference pattern remains
sufficiently high over the operating range of the instrument. A mask of this
type could be fabricated as a variable-period multilayer consisting of
alternating high-Z/low-Z materials and operated in transmission.
c) Photon flux. Estimates indicate that a detector that could record the fringe
pattern would also be feasible, even at 1.5 Å. This is based on two factors.
First, at small angles 1 of operation the wavelength of the interference
pattern will be dilated by the factor (2(1-))-1. In practical terms, periods
in the 50 -1500 Å range can be recorded. Second is that while recording
materials that operate down to this level of resolution (e.g., PMMA, or Agdoped semiconductors) are known to require large amounts of energy per unit
area (O(1 J/cm2)), such (single-shot) exposure requirements could easily be
met by the LCLS, even far away from saturation.
d) Beam splitting. It is estimated that in order to attain a 1 in the >0.005 rad
range at 1.5 Å, grating periods of 300 Å or less will be required. Such
structures, similarly to the masks described above, could be fabricated as
multilayers and operated in transmission, a technique that has been developed
in recent years at LLNL (11,12). Blurring of the LCLS temporal structure due
to the splitter's dispersive effects could to a certain extent be mitigated by
pinhole aperturing of the incoming light. Splitting methods based on
homogeneous or perforated [4] foils operating in transmission/reflection could
also be investigated, particularly if the practical performance of the grating
splitter proves to be overly dispersion-limited.
In this context, the
development of broad-band, high-quality multilayers as alternatives to the
specular reflectors assumed here could be pursued as a means of scaling down
the length of the instrument, particularly for FTS or source-analysis
applications for which the bandwidth reduction would be acceptable.
e) Alignment. The most critical issue concerns the tolerances required on the
alignments, positions, and motions of the interferometer components.
Tolerance specifications will be determined to a large extent by the optical
components utilized in the interferometer, the type of detector, and mode of
operation of the instrument. For example, starting with the splitter, it is well-
known that the far-field diffraction pattern of a normal-incidence transmission
grating is invariant with respect to the transverse coordinates of the grating.
Moreover, while it is sensitive to the inclination of the grating away from
normal incidence, the dispersion angles vary with the cosine of the deviation
angle, making the tolerance on the deviation of this parameter for controlling
the lateral motion of the split beams to , say, the ~10-9 rad level fairly robust
(~10-100 mrad). However the following elements, the mirrors, will require
exceptionally stringent tolerances both on position and angle should, for
example, maintenance of the interference pattern's lateral position on the
detector plane to a fraction of the pattern's wavelength be required. This
requirement would be necessary if operating with a mask followed by an
intensity detector (as described above), and may well represent the limit on
the lowest attainable wavelength that a practical device could operate at.
However, these tolerances could be minimized if the detector was a resist that
recorded the interference pattern of each shot. In this case the autocorrelation
could be unfolded from the distribution and statistics of the interference
pattern, and these would be substantially less sensitive to its lateral position.
In the same context, it can be noted that the ability to control the interference
patterns' lateral positions could be considerably enhanced - even for
dispersion lengths of 1-2 meters - by replacing the mirrors with transmission
grating splitters (the one corresponding to M1 being also rotatable), which
would result in the same tolerance reduction as for the incoming beam splitter.
Here again the copious flux of the LCLS would more than compensate for the
substantially lower efficiency of the grating deflectors. Needless to say, the
duration of the LCLS recording events will be so short that questions of
tolerance on any component's motion during recording can be completely
disregarded
References
[1] Fowles, G. R., Introduction to Modern Optics, New York: Holt, Rinehart and Winston, Inc., 1975, ch.
3, pp. 33-80.
[2] Murphy, J. B., Pellegrini, C. "Introduction to the Physics of the Free Electron Laser," in Frontiers of
Particle Beams, M. Month, S. Turner, eds., Lecture Notes in Physics No. 296, H. Araki et al, eds.,
Springer-Verlag, Berlin, 1988, pp. 163-212.
[3] Tatchyn, R., Arthur, J., Boyce, R., Cremer, T., Fasso, A., Montgomery, J., Vylet, V., Walz, D., Yotam,
R., Freund, A. K., Howells, M. R., "X-ray Optics Design Studies for the 1.5-15 Å Linac Coherent Light
Source (LCLS) at the Stanford Linear Accelerator Center (SLAC)," SPIE Proceedings 3154, 174-222
(1998).
[4] Moler, E. J., Duarte, R. M., Howells, M. R., Hussain, Z., Oh, C., Spring, J., "First measurements using
the ALS soft x-ray Fourier transform spectrometer," SPIE Proceedings 3154, 117-122 (1997).
9.1.2.7. Transverse coherence
Young’s slits
9.1.3 Spontaneous radiation properties
9.1.3.1 Spectral structure of higher harmonics
The reflectivity of perfect single crystals in the angstrom wavelength region is
understood quantitatively. Thus a single crystal x-ray spectrometer is the ideal tool to
study the intensity, wavelength and detailed spectral properties of the spontaneous
radiation. Such a device is both simple in concept, a precision rotation axis and a perfect
single crystal, preferably low Z to avoid damage and heating issues. The scattered
intensity can be measured with high precision with any of several detectors that are
generically ion chambers measuring the photon intensity in a current mode.
The detailed structure of the spectrum is measured in an angular scan in the theta-two
theta mode with a wide-open detector. In this way the integrated intensity as a function
of photon energy is measured. From this the incident spectrum can be quantitatively
recovered. By using the same spectrometer and measuring the angular positions of the
harmonics at plus and minus Bragg angles one can measure within the precision of the
rotary axis and the know ledge of the d-spacing of the crystal the wavelength of the
radiation without the need for determining the zero of the rotation axis, but rather just
needing the angular difference between the observed maxima at plus and minus angles.
All of these techniques are well documented in the literature and have been in use for
perhaps 50 years to characterize x-ray sources of various sorts
9.1.3.2 Energy of higher harmonics
An ionization chamber (sometimes following a crystal spectrometer, as decribed in
section 9.1.3.1, will be used.
9.1.3.3 Intensity of higher harmonics
Jerry Hastings is providing this section
9.1.3.4 Angular distribution
The measurement of the angular distribution of the harmonics of the spontaneous
radiation requires an angular slit with micro-radian resolution. This is possible with two
perfect crystals set to diffract in the so-called plus-plus arrangement. In this arrangement
the crystal pair transmits radiation of a prescribed wavelength, determined by the crystal
d-spacing and the angle between the normals to the diffracting planes of the two crystals.
The angular acceptance of the pair is determined by the width of the diffraction profile of
the perfect crystals used. The choice of this width is made by selecting the appropriate
Bragg reflection. The mechanics required are straightforward. For the angular variation
of the crystal pair a precise tangent arm mechanism with resolution on the microradian
scale with perhaps a few milliradians of range is require. The relative settings of the two
crystals must be established on the microradian scale and should be stable over the time
of the experiment. Often it is possible to choose a monolithic two-reflection system that
eliminates the need for the precise relative alignment stage with the restriction that it
works at one wavelength.
9.1.3.5 Temporal distribution
Jerry Hastings is providing this section
9.1.4 Other common usage diagnostics
9.1.4.1 Synchronization between FEL and pump or probe laser
Independent pumps and probes will require measurement of synchronization, to better
than 100 fs. It is likely that experiments will be performed in a statistical manner,
utilizing the 120 Hz repetition rate, i.e. the synchronization timing between a pump and a
probe beam will be measured after the experiments, and the data binned accordingly.
Sub-picosecond timing measurements between the FEL pump or probe photon bunch, at
> 0.8 keV, and a probe or pump that would likely be in the sub 2 eV photon energy range,
are not available and must be developed.