Large-format, high-speed, X-ray pnCCDs combined with electron and ion
imaging spectrometers in a multipurpose chamber for exper iments at 4th
generation light sources
Lothar Str
¨
uder
a,c,e,h
, Sascha Epp
a
, Daniel Rolles
a
, Robert Hartmann
b,e
, Peter Holl
b,e
, Gerhard Lutz
b,e
,
Heike Soltau
b,e
, Rouven Eckart
b,e
, Christian Reich
b,e
, Klaus Heinzinger
b,e
, Christian Thamm
b,e
,
Artem Rudenko
a
, Faton Krasniqi
a
, Kai-Uwe K
¨
uhnel
i
, Christian Bauer
i
, Claus-Dieter Schr
¨
oter
i
,
Robert Moshammer
a,i
, Simone Techert
a,m
, Danilo Miessner
c,e
, Matteo Porro
c,e
, Olaf H
¨
alker
c,e
,
Norbert Meidinger
c,e
, Nils Kimmel
c,e,
, Robert Andritschke
c,e
, Florian Schopper
c,e
,
Georg Weidenspointner
c,e
, Alexander Ziegler
c,e
, Daniel Pietschner
c,e
, Sven Herrmann
c,e
,
Ullrich Pietsch
h
, Albert Walenta
h
, Wolfram Leitenberger
h
, Christoph Bostedt
f
, Thomas M
¨
oller
f
,
Daniela Rupp
f
, Marcus Adolph
f
, Heinz Graafsma
g
, Helmut Hirsemann
g
, Klaus G
¨
artner
k
,
Rainer Richter
d,e
, Lutz Foucar
a
, Robert L. Shoeman
l
, Ilme Schlichting
a,l
, Joachim Ullrich
a,i
a
Max Planck Advanced Study Group, Center for Free Electron Laser Science (CFEL), Notkestr. 85, D-22607 Hamburg, Germany
b
PNSensor GmbH, R
¨
omerstraße 28, D-80803 M
¨
unchen, Germany
c
Max-Planck-Institut f
¨
ur extraterrestrische Physik, Giessenbachstraße, D-85741 Garching, Germany
d
Max-Planck-Institut f
¨
ur Physik, F
¨
ohringer Ring 6, D-80805 M
¨
unchen, Germany
e
MPI Halbleiterlabor, Otto-Hahn-Ring 6, D-81739 M
¨
unchen, Germany
f
Technische Universit
¨
at Berlin, Institut f
¨
ur Optik und Atomare Physik, Hardenbergstraße 36, D-10623 Berlin, Germany
g
DESY, Notkestr.85, D-22607 Hamburg, Germany
h
University of Siegen, Emmy-Noether Campus, Walter Flex Str. 3, D-57068 Siegen, Germany
i
Max-Planck-Institut f
¨
ur Kernphysik, Saupfercheckweg 1, D-69117 Heidelberg, Germany
k
Weierstrass Institut, Berlin, Mohrenstr. 39, D-10117 Berlin, Germany
l
Max-Planck-Institut f
¨
ur medizinische Forschung, Jahnstr. 29, D-69120 Heidelberg, Germany
m
Max-Planck-Institut f
¨
ur biophysikalische Chemie, Am Fassberg 11, 37077 G
¨
ottingen, Germany
article info
Article history:
Received 21 July 2009
Received in revised form
28 October 2009
Accepted 18 December 2009
Available online 4 January 2010
Keywords:
PnCCD
Parallel readout
Full depletion
Back illuminated
Radiation hard
Visible light
UV light and X-ray detection
CAMP chamber
Reaction microscope
Velocity map imaging
X-ray imaging
X-ray spectroscopy
Free electron laser
abstract
Fourth generation accelerator-based light sources, such as VUV and X-ray Free Elect ron Lasers (FEL),
deliver ultra-brilliant (10
12
–10
13
photons per bunch) coherent radiation in femtosecond (10–
100 fs) pulses and, thus, require novel focal plane instrumentation in order to fully exploit their unique
capabilities. As an additional challenge for detection devices, existing (FLASH, Hamburg) and future
FELs (LCLS, Menlo Park; SCSS, Hyogo and the European XFEL, Hamburg) cover a broad range of photon
energies from the EUV to the X-ray regime with significantly different bandwidths and pulse structures
reaching up to MHz micro-bunch repetition rates. Moreover, hundreds up to trillions of fragment
particles, ions, electrons or scattered photons can emerge when a single light flash impinges on matter
with intensities up to 10
22
W/cm
2
.
In order to meet these challenges, the Max Planck Advanced Study Group (ASG) within the Center
for Free Electron Laser Science (CFEL) has designed the CFEL-ASG MultiPurpose (CAMP) chamber. It is
equipped with specially developed photon and charged particle detection devices dedicated to cover
large solid-angles. A variety of different targets are supported, such as atomic, (aligned) molecular and
cluster jets, particle injectors for bio-samples or fixed target arrangements. CAMP houses 4
p
solid-angle
ion and electron momentum imaging spectrometers (‘‘reaction microscope’’, REMI, or ‘‘velocity map
imaging’’, VMI) in a unique combination with novel, large-area, broadband (50 eV–25 keV), high-
dynamic-range, single-photon-counting and imaging X-ray detectors based on the pnCCDs.
This instrumentation allows a new class of coheren t diffraction experiments in which both electron
and ion emission from the target may be simultaneously monitored. This permits the investigation of
ARTICLE IN PRESS
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/nima
Nuclear Instruments and Methods in
Physics Research A
0168-9002/$ - see front matter & 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.nima.2009.12.053
Corresponding author at: Max-Planck-Institut f
¨
ur extraterrestrische Physik,
Giessenbachstraße, D-85741 Garching, Germany.
E-mail addresses: nik@hll.mpg.de, nils.kimmel@hll.mpg.de (N. Kimmel).
Nuclear Instruments and Methods in Physics Research A 614 (2010) 483–496
ARTICLE IN PRESS
dynamic processes in this new regime of ultra-intense, high-ener gy ra diation—matter interaction. After
an introd uction into the salient features of the CAMP chamber and the properties of the redesigned
REMI/VMI spectrometers, the new 1024 1024 pixel format pnCCD imaging detector system will be
described in detail. Results of tests of four smaller format (256 512) devices of identical performance,
conducted at FLASH and BESSY, will be presented and the concept as well as the anticipated properties
of the full, large-scale system will be elucidated. The data obtained at both radiation sources illustrate
the unprecedented performance of the X-ray detectors, which have a voxel size of 75 75 450
m
m
3
and a typical read-out noise of 2.5 electrons (rms) at an operating temperat ure of 50 1C.
& 2010 Elsevier B.V. All rights reserved.
1. Introduction
Since the beginning of this decade, several laboratories
worldwide have decided to build a new generation of extremely
intense, coherent and short-pulsed VUV and X-ray light sources:
the (X-ray) free electron lasers or (X-)FELs. These revolutionary
machines deliver up to 10
13
photons in 10–100 fs bursts, in small,
parallel beams with an energy uncertainty of
D
E/E10
3
. They
thus reach peak brilliances up to nine orders of magnitude higher
than those achieved at the most advanced third-generation
synchrotrons. Due to the small source size, the beam can be
focused down to submicron spots, giving light intensities in the
focus of up to 10
22
W/cm
2
. The first machine of this kind was the
Free electron LASer at Hamburg, FLASH at DESY [1]. FLASH has
been operational since 2005 and produces light at wavelengths
between 6 and 40 nm (i.e., 30 up to 200 eV) and pulse durations of
about 25 fs. The next large facility to come online and presently in
its commissioning phase is the Linac Coherent Light Source (LCLS)
at SLAC in Menlo Park. LCLS is designed to deliver X-rays from
below 0.8 up to 8 keV at 100 fs pulse duration [2]. While their
basic concepts are identical, the practical realization and the
operational schemes of FLASH and LCLS are quite different. FLASH
employs super-conducting accelerator cavities with a fundamen-
tal repetition rate of 5 or 10 Hz and can optionally deliver up to
3000 (design value) micro-bunches per macro-pulse train, while
LCLS will operate with conventional accelerator cavities at a
constant repetition rate between 30 and 120 Hz. In 2010, the
SPring8 Compact SASE Source (SCSS) will be commissioned in
Japan [3] with beam parameters similar to LCLS and a repetition
rate of 60 Hz. Finally, in 2014, another new facility will be opened
to the scientific public, the X-ray Free Electron Laser, the European
XFEL, in Hamburg [4]. The European XFEL will have a photon
bunch structure similar to the one at FLASH. Every 100 ms, 3000
photon pulses, equally spaced in time by 200 ns, will be
generated. After 600
m
s, there will be a pause of 99.4 ms, which
is the cooling phase of the cryogenic super-conducting accelerator
cavities. This means that on average 30,000 X-ray pulses will be
delivered per second, providing 250 times the mean luminosity of
LCLS and about 300 times the total photon flux achieved at PETRA
III at DESY, the most advanced synchrotron in this energy range.
Accordingly, it is envisioned that major breakthroughs in
different scientific disciplines ranging from physics and chemistry
to material and life sciences will be achieved with these discovery
machines. Foremost among them is the dream to follow the
‘‘making and breaking’’ of molecular bonds in real-time, i.e., to
record a ‘‘molecular movie’’. In biology, one hopes to obtain
structural information with sub-nanometer resolution of materi-
als beyond the scope of conventional crystallographic approaches.
These include nano- or non-crystalline proteins, macromolecular
assemblies, viruses and cells. If successful, this approach would
allow another access to the structure of membrane proteins
which constitute about 60% of all therapeutic drug targets and are
notoriously difficult to crystallize. In general, due to their high
coherence, brilliance and short pulse time properties, FELs allow
structural information to be obtained as a function of time, which
will allow research in the femtosecond time regime. Beyond
applications in biology and chemistry, this research is expected to
be of utmost importance for the understanding of strongly
correlated or magnetic materials, of phase transitions in liquids
and solids, of melting processes at surfaces or in the bulk, of
electron dynamics in atoms and molecules and of strongly
correlated plasmas.
While these new machines promise unprecedented scientific
possibilities, their properties also pose extreme challenges to the
experimental instrumentation. For example, in some cases,
scattered photon images have to be recorded at up to 5 MHz frame
readout rates or, in other cases, hundreds, thousands or even more
electrons and ions have to be detected per shot. Beyond the
requirements with respect to the performance of the detectors
themselves, petabytes of data are recorded and have to be analyzed
within reasonable times. Moreover, fundamental scientific ques-
tions arise beyond the ‘‘mere’’ technological challenges when
penetrating this true terra incognita of light-matter interaction,
which is orders of magnitude away from our present experience in
energy density at the given photon energies. Thus, one of the
decisive open questions regarding coherent imaging of biomole-
cules in the gas phase is whether the objects can be imaged before
they are destroyed in the super-intense light flash. Since each
single shot image will be a 2 dimensional projection of a 3
dimensional object, it will be necessary to know or to calculate the
relative orientation of the molecule for each single shot before
calculating a 3 dimensional reconstruction.
In order to meet some of these challenges concerning the
instrumentation and to explore the relevant many-particle
dynamics in this uncharted territory, we have designed the
CFEL-ASG Multipurpose (CAMP) chamber. Proposed at the end of
2007 and designed in 2008 after extensive discussions within the
scientific community, this chamber was built, tested and operated
in 2009. The CAMP chamber houses the world’s largest pnCCD
chips. These detectors operate with a frame readout rate of up to
200 Hz and are thus fully capable of meeting the requirements at
LCLS (up to 120 Hz) and SCSS (up to 60 Hz), as well as the 10 Hz
macro-bunch operation scheme at FLASH or the European XFEL.
The high-speed European XFEL and FLASH operation at up to
5 MHz is expected to be coped with the recently proposed DePFET
active pixel sensors and hybrid pixel detector arrays [5] .In
addition, large-solid-angle momentum imaging spectrometers for
emitted electrons and ions, which are commonly referred to as
‘‘reaction microscopes’’ (REMI) [6] or ‘‘velocity map imaging’’
(VMI) systems [7], have been redesigned and upgraded in order to
enable simultaneous operation with the pnCCDs and to accom-
modate the fact that hundreds or even thousands of electrons and
ions are emitted per shot. This combination of pnCCD photon
detection with advanced charged-particle detection is expected to
deliver detailed information on the interaction dynamics of
intense VUV and X-ray FEL pulses with matter, to help uncover
and subsequently possibly modify the destruction mechanisms of
bio-particles to be imaged, to find schemes to dynamically align
molecules for different shots by inspecting their fragmentation
and scattering patterns shot by shot, etc.
L. Str
¨
uder et al. / Nuclear Instruments and Methods in Physics Research A 614 (2010) 483–496484
ARTICLE IN PRESS
Thus, we envision that the CAMP chamber will serve a broad
community encompassing coherent X-ray scattering and lensless
imaging in material science (CXI), single particle and biomolecule
imaging (SPB), femtosecond diffraction imaging experiments (FDE),
X-ray photon correlation spectroscopy (XPC), high energy density
matter (HED) research, atomic and molecular physics (AMO), science
with small quantum systems (SQS), X-ray absorption spectroscopy
(XAS) and many more research areas to come. The unique
combination of cutting edge imaging detectors along with the
multi-user capability of the design was impressively affirmed by the
fact that about 30% of all approved (first campaign until December
2009) and proposed (second campaign starting in March 2010)
experiments at LCLS plan to make use of the CAMP chamber.
Moreover, a partial copy of the CAMP setup is presently under
development for the Japanese SCSS free electron laser and has been
designed such that our pnCCD detector array including chambers C2
and C3 (see Fig. 1) can be directly implemented. Experiments in
Japan are planned to be performed in 2010.
This paper describes the general concept of the CAMP
chamber, the REMI/VMI setup and the X-ray imagers. Since the
REMI and VMI concepts have been discussed previously [6],we
will concentrate on the changes and improvements necessary to
combine them, for the first time, with photon detection. We will
then focus on the detailed description of the pnCCD detector
concept and elucidate how the pnCCD will meet the performance
requests put forward by the users of the CAMP chamber. We will
discuss the physical limitations of the system in terms of readout
speed, noise performance, quantum efficiency, dynamic range,
charge handling capacity, the dynamics of the charge cloud
expansion and long term stability. We will further report on
relevant previous results based on experiments performed at the
BESSY synchrotron and at the FLASH facility. In this experiments,
pnCCDs with an identical geometry (voxel size: 75 75 450
m
m
3
with a format of 256 256 2) were tested at a wide range
of energies (30 eV–35 keV) and intensities and provide a basis for
extrapolation to the anticipated performance of the pnCCD
detectors in the full CAMP chamber.
2. The CFEL-ASG Multipurpose (CAMP) chamber
As outlined before, the CAMP setup is designed to achieve two
major goals:
(i) To obtain the most complete data sets possible by momen-
tum-resolved, large-solid-angle correlated detection of a large
number of the electrons, ions and scattered or fluorescent
photons emitted in each shot, similarly to the data collection
strategies employed in high-energy physics experiments. This
shall be achieved for different situations in which the FEL
beam hits atoms, molecules, solids or plasmas or when optical
alignment as well as pump lasers are fed in to allow for pump-
probe experiments on aligned molecules.
(ii) To guarantee high flexibility in order to enable the execution
of a large variety of experiments. The ion and electron
spectrometers allow for (coincident) many-particle, ion and
electron momentum spectroscopy, as well as for velocity map
imaging in combination with various jet targets. These
components can be readily removed and replaced by other
devices, such as fixed target setups, particle injectors or high-
resolution electron time-of-flight (TOF) spectrometers. More-
over, as will be illustrated below, the photon imaging
detectors can even be mounted on the side of the incoming
beam for more efficient detection of back-reflected or
fluorescence light.
2.1. General concept, layout and dimensions
Fig. 1 depicts the complete CAMP setup including the support
and the Helmholtz coils for generating the magnetic projection
field for the electrons in the REMI. The FEL beam enters the
chamber from the left and can exit towards a beam dump to the
right. The device consists of four sections, C0–C3. From left to
right, it starts with a differential pumping stage (C0), which
Fig. 1. (a) Side-view and dimensions of the CAMP chamber, which consists of the four chamber sections C0–C3. The FEL beam enters from the left and is focused onto the
target at the interaction point at the center of C1. The whole chamber assembly can be remotely adjusted via translational motions of the inner support (greenish brown)
along each of the three spatial directions and tilt motions along the vertical and horizontal angles around the pivotal point (see text). The base support (blue) remains fixed.
(b) Three-dimensional view of the CAMP chamber assembly visualizing the Helmholtz coils that generate a magnetic field of up to 70 Gauss for momentum imaging of
electrons in the REMI. As an example of one hardware configuration, a supersonic jet target (atoms, molecules, clusters) is illustrated entering from the left along the y
direction, with the jet dump on the opposite side. Large, removable flanges permit simple hardware exchanges.
L. Str
¨
uder et al. / Nuclear Instruments and Methods in Physics Research A 614 (2010) 483–496 485
ARTICLE IN PRESS
ensures that the beamline vacuum requirements are met even for
main chamber pressures up to 10
6
mbar as may be present for
injection of dense target beams. The first main chamber (C1),
which has the FEL-sample interaction point at its center,
houses the REMI or VMI spectrometers in the vertical direction
(see Fig. 2a), and the jet-target, particle injectors or fixed-target
setups. The first movable pnCCD detector array, pnCCD1, is
located in the next large chamber (C2) whereas the second
detector array, pnCCD2, is accommodated in C3 (see Fig. 2b). C1
and C2 have a +400 mm clear bore along the laser beam
direction, whereas C3 has a +250 mm clear bore.
The design base pressure in the whole chamber is typically
around 10
8
mbar, but can be lowered to 10
10
mbar if required
for particular experiments. To be able to reach this ultra-high
vacuum range, flexible and easy-to-mount helicoflex gaskets are
used for the 400 mm flanges. To allow for remote three-
dimensional position and angular alignment, which is crucial for
the proper operation of the pnCCDs, the whole chamber assembly
is mounted on an inner support (greenish in Fig. 1) which can be
moved independently along the x, y and z directions by 750,
750 and 7150 mm, respectively, with respect to the main
support (blue in Fig. 1), which remains fixed. Thus, the target
point can be adjusted exactly to the FEL beam. Thereafter, the
center of pnCCD2, which is fixed within C3, can be aligned with
respect to the FEL beam direction by rotating the whole assembly
in the horizontal as well as vertical directions around a fixed
pivotal point, each by an angle of about 7 61 as depicted in Fig. 1a.
Since this pivotal point does not coincide with the target location
for technical reasons, an iterative procedure with subsequent 3D
and angular alignment actions may be necessary in some cases.
The unique and new feature provided by the fully equipped
endstation is the possibility to simultaneously operate electron,
ion and photon momentum imaging detectors as illustrated in
cross sections through C1 in Fig. 2a, as well as through C1–C3 in
Fig. 2b. Conically shaped electrodes around the interaction point
allow for a free line of sight to the interaction point from all
directions and generate homogeneous (REMI) or inhomogeneous
(VMI) electric fields to extract ions and electrons into different
directions towards two-dimensional position and time-resolved
multi-channel detectors. In the standard configuration, photons
are detected via two pnCCD detector arrays placed downstream of
the interaction point as depicted in Fig. 2b. In this figure, the
REMI/VMI was removed for better illustration of the pnCCDs, and,
as an example, a supersonic jet source is shown in combination
with a fed-in laser that allows aligning the target molecules. We
would like to stress that the full REMI/VMI setup can be included
in a way that is fully compatible with the full position
maneuverability of the pnCCD system.
Two full sets of imaging X-ray detectors, each with ca.
80 80 mm
2
, 1024 1024 pixels active detection area (Figs. 2, 7
and 9) are incorporated into the CAMP chamber. Both detector
systems consist of two halves with 1024 512 pixels. They are
mounted on an electro-mechanical sub-module (see also Fig. 10).
In total, four identical sub-modules will be installed in the CAMP
chamber. The first detector set (pnCCD1) is moveable (Figs. 2 and
4) along the beam direction over 250 mm with the closest position
being 50 mm behind the focal point at the center of C1. The two
halves of the first detector can each be moved vertically (i.e.
perpendicular to the beam) by 45 mm in order to cover larger
scattering angles and can also overlap to shrink the diameter of
the central hole if needed. The sketch in the right part of Fig. 2
illustrates possible motions of pnCCD1. The second set, pnCCD2, is
housed in C3 and attached via a CF-DN250 flange at C2 in a fixed
geometry with a minimal distance of 500 mm from the intersec-
tion point (standard position), but can be moved to any longer
distance (by inserting adaptor tubes). Either one or both of the
two detectors sets can also be used in back-scattering geometry
by just feeding in the FEL beam in reversed, x direction or by
mounting either pnCCD2 or pnCCD1 at the upstream side of C1
with the incoming photon beam traversing the detectors through
the center hole of +2.2 mm. This photon detector geometry is
ideally suited to back-reflect low-energy photon beams via
multilayer mirrors in order to achieve smallest focus spots
and, thus, highest intensities as recently demonstrated at FLASH
(see e.g. [8]). We would like to further point out that mounting
additional detector sets in the upstream direction is one natural,
upcoming extension of CAMP in the future (Figs. 3–10).
2.2. The target chamber and the reaction microscope setup
The REMI/VMI spectrometer (outer +90 mm for VMI and
135 mm for REMI, shown in yellow in left part of Figs. 2), is
mounted vertically in the standard setup such that it does not
Fig. 2. Schematic section through the CAMP chambers C1 (left) and C1–C3 (right). The reaction microscope (REMI) or velocity map imaging (VMI) setup with specially
designed electrodes is depicted on the left hand side. The conically shaped electrodes guarantee a free line of sight into the interaction target volume from all directions.
Thus, the pnCCD detectors (for their 3D positions, see the right hand side, drawn without the REMI/VMI for illustration purposes) can detect photons, emerging under
almost any angle from the target point. In addition, this design allows feeding in other lasers for alignment or pump-probe purposes, as well as for mounting other high-
resolution, small-solid-angle electron TOF or crystal spectrometers. The pnCCD1 close to the REMI can be moved in all three directions with a maximum distance of 25 cm
along the beam trajectory. The second pnCCD2 is fixed on the flange and can be aligned with respect to the FEL beam as illustrated in Fig. 1a by the procedure described in
the text.
L. Str
¨
uder et al. / Nuclear Instruments and Methods in Physics Research A 614 (2010) 483–496486
ARTICLE IN PRESS
limit the motion of pnCCD1. This permits the mounting of
different target jet devices along with lasers for alignment or
pump-probe purposes. The middle part of each spectrometer
consists of conically shaped, gold-plated aluminum electrodes
that generate the homogeneous projecting (REMI, Fig. 3a) or
inhomogeneous imaging (VMI, Fig. 3b) electric fields. Further
outside, two solid drift tubes (VMI) or equally spaced potential
rings (REMI) guarantee a field-free drift region. The conical shape
of the central electrodes within the solid angle covered by the
pnCCD detectors allows unrestricted view into the scattering
volume. At the same time, the cones excellently suppress
scattered stray radiation emerging from outside the interaction
center.
Since the operation principles of REMI/VMI spectrometers
have been described extensively in the literature, only the salient
features are briefly outlined in the present contribution. For the
REMI configuration, illustrated in Figs. 2a and 3a, the electric field
forces all ions, independent of their initial momentum, onto
parabolic trajectories (red lines in Fig. 2a) and projects them onto
the detector. The electric field, usually a few V/cm, can in most
cases be adjusted such that all ions of interest are registered.
Measuring the time-of-flight (TOF) and the hit position, the
trajectory and, thus, the full three dimensional momentum
vector of each individual ion can be reconstructed. Electrons are
pushed towards the other hemisphere by the electric field but
cannot be efficiently projected onto the detector with moderate
fields. Even though their momenta are typically very similar to
those of the ions, the electron energies are several thousand times
larger, and accordingly, keV projection voltages would be
required, leading to times-of flight of a few nanoseconds. This is
too short to achieve reasonable momentum resolution along the
extraction field direction. In a REMI, the dilemma is solved by
maintaining a small electric field but confining the electrons’
transverse motion via an additional homogeneous (Helmholtz
configuration, coil +160 cm) magnetic field (here up to 70
Gauss), oriented parallel to the electric field and forcing them
onto spiral trajectories as sketched in Fig. 2a (green line). Again,
by measuring the TOF and the hit position, the full initial
momentum vector can be retrieved (for details, such as the
attainable resolution, see [6]). Both detectors are equipped with
either square or hexagonal delay line anodes (Roentdek HEX80
[9]) read out by an eight-channel, 2 GHz digitizer (Acquiris
DC82 2) recording the complete wave form of the signal. It
has been demonstrated recently at the SCSS test facility that up to
150 ions emerging from the interaction of a single EUV-FEL pulse
with a single cluster could be analyzed, with their individual 3D
momentum vectors [10] determined independently. In principle,
the REMI allows for the coincident detection of momentum-
resolved ions and electrons emerging from the same reaction. This
feature has been used to explore sequential two-photon double
ionization of Ne atoms in a kinematically complete experiment
where two electrons and the ion have been detected in a triple
coincidence [11]. Meanwhile, specially adapted REMIs have been
successfully operated in many experiments at FLASH and the SCSS
test facility [12–14] and the present design builds on this previous
experience.
Alternatively, if even more fragments are ejected per FEL shot,
the delay line anode can be replaced by a phosphor screen
combined with a CCD camera readout, thus recording the image
of electrons and/or ions without the magnetic field and applying
non-homogeneous imaging voltages generated by the potential
cones depicted in Fig. 3b. In this ‘‘velocity map imaging’’ [7] mode,
only two-dimensional (position) images are generated and
the time information is usually lost due to short flight times of
the electrons resulting from the large imaging electric fields of
up to keV/cm. Still, under certain conditions and/or for
low-energy electrons, three-dimensional images can be retrieved
(for details see [15]). Again, the applicability of this method at
FELs has been proven recently at FLASH [16]. Since the standard
VMI electrode configuration [7] had
to be modified for our
application, additional lenses are used to provide better space
focusing [17].
Fig. 4. Artist’s view of the two pnCCD mountings. The two halves of the pnCCD on
the left side (closest to the interaction point), CCD1, can be opened vertically by up
to 90, 45 mm each. Photons scattered under large angles can hence be detected,
while the second (fixed) pnCCD2 records the X-rays under smaller angles. The
distance between the detectors can be changed by moving the CCD1 system up to
250 mm along the beam axis. Moreover, if needed, CCD2 can be placed at any
distance larger than 550 mm away from the interaction zone by simply
introducing fixed spacer tubes of the desired length.
Fig. 3. 3D model of the CAMP reaction microscope (a) and the velocity map
imaging (b) spectrometers. The total length of both devices is 400 mm, and both
are designed for +80 mm detectors. The conically shaped central electrodes are
clearly visible. They provide a free line of sight for the pnCCDs over their complete
active area and, at the same time, suppress stray light, which is, especially effective
for the REMI geometry. For the REMI typical voltages are a few V/cm. The VMI is
operated at kV/cm for full solid-angle imaging of electrons up to 200 eV. With a
magnetic field of up to 70 Gauss for the REMI, electrons emitted with energies of
up to 2 keV into the most unfavourable direction, transverse to the electric
extraction field, are confined to a spiral trajectory with a diameter of about 70 mm
and, thus, are detected with a solid angle of 50% of the full 4
p
steradiant. Multi-
channel plate detectors are mounted at both ends and can be equipped with delay-
line anodes or phosphor screen CCD camera readout (see text).
L. Str
¨
uder et al. / Nuclear Instruments and Methods in Physics Research A 614 (2010) 483–496 487
![Fig. 15. Energy line scan along the specular direction extracted from the 3-D data shown in Fig. 14. Data from the pnCCD (filled circles) are compared to the results obtained with an energy dispersive point detector (line denoted by ‘‘SDD’’; see text). The detector systems were not cross calibrated. For details see [22].](/figures/fig-15-energy-line-scan-along-the-specular-direction-2zmh35hd.webp)
