X-Ray Diffraction from Isolated and Strongly Aligned Gas-Phase Molecules
with a Free-Electron Laser
Jochen Küpper,
1,2,3,4,5,*
Stephan Stern,
1,2
Lotte Holmegaard,
1,6
Frank Filsinger,
4,5,a
Arnaud Rouzée,
7,8
Artem Rudenko,
5,9,10
Per Johnsson,
11
Andrew V. Martin,
1,b
Marcus Adolph,
12
Andrew Aquila,
1,21
Saša Bajt,
21
Anton Barty,
1
Christoph Bostedt,
13
John Bozek,
13
Carl Caleman,
1,14
Ryan Coffee,
13
Nicola Coppola,
1
Tjark Delmas,
1
Sascha Epp,
5,9
Benjamin Erk,
5,9,c
Lutz
Foucar,
5,15
Tais Gorkhover,
12
Lars Gumprecht,
1
Andreas Hartmann,
16
Robert Hartmann,
16
Günter Hauser,
17,18
Peter Holl,
16
Andre Hömke,
5,9
Nils Kimmel,
17
Faton Krasniqi,
5,15
Kai-Uwe Kühnel,
9
Jochen Maurer,
6
Marc Messerschmidt,
13
Robert
Moshammer,
9,5
Christian Reich,
16
Benedikt Rudek,
5,9,d
Robin Santra,
1,2,3
Ilme Schlichting,
15,5
Carlo Schmidt,
5
Sebastian
Schorb,
12
Joachim Schulz,
1,e
Heike Soltau,
16
John C. H. Spence,
19
Dmitri Starodub,
19,f
Lothar Strüder,
17,20,g
Jan Thøgersen,
6
Marc J. J. Vrakking,
7,8
Georg Weidenspointner,
17,18
Thomas A. White,
1
Cornelia Wunderer,
21
Gerard Meijer,
4,h
Joachim
Ullrich,
9,5,d
Henrik Stapelfeldt,
6,22
Daniel Rolles,
5,15,21
and Henry N. Chapman
1,2,3
1
Center for Free-Electron-Laser Science (CFEL), DESY, Notkestrasse 85, 22607 Hamburg, Germany
2
Department of Physics, University of Hamburg, Luruper Chaussee 149, 22761 Hamburg, Germany
3
Hamburg Center for Ultrafast Imaging, University of Hamburg, Luruper Chaussee 149, 22761 Hamburg, Germany
4
Fritz Haber Institute of the MPG, Faradayweg 4–6, 14195 Berlin, Germany
5
Max Planck Advanced Study Group at CFEL, Notkestrasse 85, 22607 Hamburg, Germany
6
Department of Chemistry, Aarhus University, 8000 Aarhus C, Denmark
7
FOM Institute AMOLF, Science Park 104, 1098 XG Amsterdam , Netherlands
8
Max-Born-Institute, Max Born Strasse. 2a, 12489 Berlin, Germany
9
Max Planck Institute for Nuclear Physics, 69117 Heidelberg, Germany
10
J. R. Macdonald Laboratory, Department of Physics, Kansas State University, Manhatt an, Kansas 66506, USA
11
Department of Physics, Lund University, P. O. Box 118, 22100 Lund, Sweden
12
Technical University of Berlin, 10623 Berlin, Germany
13
Linac Coherent Light Source, SLAC National Accelerator Laboratory, 2575 Sand Hill Road, Menlo Park, California 94025, USA
14
Uppsala University, Department of Physics and Astronomy, Box 516, 75120 Uppsala, Sweden
15
Max Planck Institute for Medical Research, 69120 Heidelberg, Germany
16
PNSensor GmbH, 81739 Munich, Germany
17
Max Planck Semiconduc tor Laboratory, 81739 Munich, Germany
18
Max Planck Institute for Extraterrestrial Physics, 85741 Garching, Germany
19
Department of Physics, Arizona State University, Tempe, Arizona 85287, USA
20
University of Siegen, Emmy-Noether Campus, Walter Flex Str. 3, 57068 Siegen, Germany
21
Deutsches Elektronen-Synchrotron (DESY), 22607 Hamburg, Germany
22
Interdisciplinary Nanoscience Center (iNANO), Aarhus University,
8000 Aarhus C, Denmark
(Received 16 July 2013; published 28 February 2014)
We report experimental results on x-ray diffraction of quantum-state-selected and strongly aligned
ensembles of the prototypical asymmetric rotor molecule 2,5-diiodobenzonitrile using the Linac Coherent
Light Source. The experiments demonstrate first steps toward a new approach to diffractive imaging of
distinct structures of individual, isolated gas-phase molecules. We confirm several key ingredients of single
molecule diffraction experiments: the abilities to detect and count individual scattered x-ray photons in
single shot diffraction data, to deliver state-selected, e.g., structural-isomer-selected, ensembles of
molecules to the x-ray interaction volume, and to strongly align the scattering molecules. Our approach,
using ultrashort x-ray pulses, is suitable to study ultrafast dynamics of isolated molecules.
DOI: 10.1103/PhysRevLett.112.083002 PACS numbers: 33.15.-e, 33.15.Dj, 33.80.-b, 37.10.-x
X-ray free-electron lasers (XFELs) hold the promise for
determining atomically resolved structures and for tracing
structural dynamics of individual molecules and nano-
particles [1] . Over the last decade, ground-breaking experi-
ments were performed at the Free-Electron Laser in
Hamburg (FLASH) at DESY [2–5] and the Linac
Coherent Light Source (LCLS) at the SLAC National
Accelerator Laboratory [6–12]. These experiments already
begin to provide new insights into fundamental aspects of
matter, such as hitherto unobserved structures of non-
crystallizable mesoscopic objects [13–15] or the radiation
damage induced by the short and very strong x-ray pulses
[8,16]. However, the path to actual determination of
atomically resolved structures and dynamics of single
PRL 112, 083002 (2014)
PHYSICAL REVIEW LETTERS
week ending
28 FEBRUARY 2014
0031-9007=14=112(8)=083002(6) 083002-1 © 2014 American Physical Society
molecules is still long [13]. Nevertheless, related experi-
ments on the investigation of small-molecule structures and
their dynamics utilizing molecular ensembles are within
reach [13,17].
To be able to record structural changes during ultrafast
molecular processes under well-defined conditions, it was
proposed [13,17] to spatially separate shapes [18],sizes
[19],orindividualisomers[20–22] of complex small
molecules before delivery to the interaction point of an
XFEL. The molecules should be one- or three-dimensionally
aligned or oriented in space [11,12,17,23–28].Thiscon-
trolled-delivery approach would allow for the averaging of
many identical patterns, similar to recent electron diffraction
experiments on aligned CF
3
I [28] or to photoelectron
imaging of 1-Ethynyl-4-fluorobenzene [12]. A controlled
variation of the alignment direction in space allows one to
tomographically build up the complete three-dimensional
diffraction volume of individual isomers. This ensemble-
and pulse-averaging approach would allow working at
appropriately low fluences to circumvent detrimental elec-
tronic damage processes that have been predicted [29–31]
for the very high x-ray fluences necessary to obtain
classifiable single-molecule diffraction patterns. The forth-
coming European XFEL facility will give the opportunity to
collect patterns at a rate of 27 000 per second, which could
be sufficient to collect the necessary 10
5
–10
8
patterns within
minutes or hours [13].
Here, we record x-ray-diffraction patterns of
ensembles of identical, state-selected, and strongly aligned
2,5-diiodo-benzonitrile (DIBN, Fig. 1) molecules in the
gas phase, demonstrating the applicability of this
controlled-delivery approach. Using 2 keV (620 pm) radi-
ation from the LCLS, we succeeded in observing the two-
center interference between the two iodine scattering
centers, separated by approximately 700 pm, in the
continuous coherent diffraction pattern. The strongly
aligned samples [32] allow us to simply average the
continuous diffraction patterns from a very large number
of isolated molecules [13,17]. We restricted the angular
control to one-dimensional alignment of the axis containing
the two iodine atoms, as this was the solely required control
for this experiment. The extension to three-dimensional
alignment and orientation is straightforward for the cold,
state-selected samples employed [33–35]. Moreover, we
have previously demonstrated that, for more complex
molecules, we could also exploit the current setup to
spatially separate structural isomers and sizes [19–21].
The experiment was performed at the AMO beamline at
LCLS [6,7] using the CAMP endstation [36,37] extended
by a state-of-the-art molecular beam setup [38]. Figure 1
shows a scheme of the experimental arrangement. The
setup contains multiple devices to simultaneously detect
photons, electrons, and ions [36]. A pulsed cold molecular
beam is formed by expanding a few mbar of DIBN in
50 bar of helium into a vacuum through an Even-Lavie
valve [39]. The molecular beam travels through an electro-
static deflector, which disperses the molecules according to
their rotational quantum states, into the target region.
There, it is crossed by three pulsed laser beams: One laser
beam consisting of 12 ns (FWHM) pulses from a Nd∶YAG
laser (YAG, λ ¼ 1064 nm, E
I
¼ 200 mJ, ω
0
¼ 63 μm,
I
0
≈ 2.5 × 10
11
W=cm
2
) is used to align the molecules.
A second laser beam consists of 60 fs (FWHM) pulses from
a Ti:sapphire laser (TSL, 800 nm, E
I
¼ 400 μJ,
ω
0
¼ 40 μm, I
0
≈ 2.5 × 10
14
W=cm
2
) and is used to opti-
mize the molecular beam and the alignment without LCLS.
The third beam consists of the 100 fs x-ray pulses
(LCLS, λ ¼ 620 pm (2 keV), E
I
¼ 4 mJ, ω ¼ 30 μm,
I
0
≈ 2 × 10
15
W=cm
2
); we estimate that 35% of the gen-
erated 1 .25 × 10
13
x-ray photons/pulse are transported to
the experiment [40]. All three laser beams are copropagat-
ing, overlapped using dichroic (1064 and 800 nm) and
holey [near infrared (NIR) lasers and x rays] mirrors before
they intersect the sample and finally leave the setup through
a gap in an on-axis pnCCD camera and another holey
mirror to separate the laser beams again. Time-of-flight and
velocity-map-imaging (VMI) spectrometers are installed
perpendicular to the horizontal plane of the molecular and
laser beams to investigate the ion- and electron-momentum
distributions resulting from the Coulomb explosion due to
absorption of one or a few x-ray photons.
We exploit Coulomb explosion imaging of DIBN
induced either by strong field ionization using the TSL
pulse or through one- or two-photon ionization by the x-ray
pulse to analyze the alignment of the rotational-state-
selected molecules along their I—I axis. The pertinent
x
y
z
High-Voltage electrodes
CCD camera
Multi-channel plate + Phosphor screen
DIBN
Holey mirror
Electrostatic
beam deflector
Even-Lavie
valve
Holey mirror
Delay line
anode
pnCCD detector
Ion velocity-map imaging
FEL
700pm
620pm
z (mm)
y (mm)
−3 −2 −1 0 1 2 3
−2
−1
0
1
2
3
+10 kV
-10 kV
E (kV/cm)
60
80
100
120
140
Light-baffling tube
FEL
YA G
+
TSL
Light-baffling tube
Skimmer
Skimmer
Molecular beam
electrostatic
deflection field
FIG. 1 (color online). Schematic view of the experimental
setup: from the left, a supersonic beam with quantum-state
selected molecules is delivered to the interaction point. In the
center of a dual velocity map imaging spectrometer, the molecu-
lar beam is crossed by laser beams copropagating from right to
left. The direct laser beams go through a gap in the pnCCD
detectors that are used to record the diffraction pattern. The upper
pnCCD panel is further away from the beam axis than the bot tom
panel in order to cover a wider range of scattering angles. In the
inset, the molecular structure of 2,5-diiodobenzonitrile is de-
picted, together with a scale of its size, i.e., the iodine-iodine
distance, and the wavelength of the x rays.
PRL 112, 083002 (2014)
PHYSICAL REVIEW LETTERS
week ending
28 FEBRUARY 2014
083002-2
experimental observable is the emission direction of the
recoiling I
þ
ions from the Coulomb explosion, illustrated
by the 2D I
þ
ion images in Fig. 2. Without the YAG pulse,
the I
þ
images (Fig. 2) were circularly symmetric as
expected for randomly aligned molecules. The circularly
symmetric image obtained following ionization with the
horizontally polarized LCLS beam demonstrated that the
interaction of the far-off resonant radiation with
the molecule was independent of the angle between the
molecular axis and the x-ray polarization direction: The
x rays were a practically unbiased ideal probe of the spatial
orientation of the molecules. When the YAG pulse was
included, the I
þ
ions were strongly confined along the YAG
polarization axis demonstrating tight adiabatic 1D align-
ment. From the corresponding 2D momentum distribution
shown in Figs. 2(b) and 2(d), we extracted hcos
2
θ
2D
i¼
0.89 and hcos
2
θ
2D
i¼0.88 for the TSL and LCLS ioniza-
tion, respectively. This degree of alignment is in good
agreement with previous measurements of adiabatic align-
ment of similar molecules [32] and stronger than previous
alignment experiments of diatomic molecules at the LCLS
[11]. This demonstrated strong alignment of complex
molecules, even under the constraint conditions of a
temporary setup at a FEL beamline. It was made possible
by the very cold molecular beam and the adiabatic align-
ment conditions. The demonstrated degree of alignment
fulfills the requirements for observing aligned molecule
diffraction [17,24,28].
In subsequent experiments we recorded the x-ray dif-
fraction data of these aligned samples on the pnCCD
cameras. For these experiments the polarization of the
YAG laser was rotated clockwise by α ¼ −60°. VMI data
were repeatedly recorded in between diffraction experi-
ments under the same conditions as in Fig. 2. An average
value for the degree of alignment in the diffraction data of
hcos
2
θ
2D
i¼0.84 was derived, limited by the (changing)
spatial overlap of the foci of the YAG and the LCLS beams.
The obtained x-ray diffraction patterns are shown in Fig. S1
in the Supplemental Material (SM) [41]. We have analyzed
diffraction data for ≈563000 shots with YAG and ≈842000
shots without (NOYAG), respectively, corresponding to 7 h
(YAG) and 9 h (NOYAG) measurement time with LCLS
operating at 60 Hz. These data have been corrected for
background and camera artifacts and individual photon hits
are extracted (see SM [41]). This results in 0.20 photons/
shot, which are placed in a histogram that represents the
molecular diffraction pattern (Fig. S2). By subtracting
the diffraction pattern of randomly oriented molecules
(I
NOYAG
) from the diffraction pattern of aligned molecules
(I
YAG
), the background is cancelled. This includes the
isotropic background originating from atomic scattering of
the atoms in the DIBN molecule and the helium seed gas, as
well as experimental background, e.g., scattering from
apertures and rest gas.
In Fig. 3, we present these diffraction differences
(I
YAG
− I
NOYAG
) for simulated [Figs. 3(a) and 3(c)] and
experimentally observed [Figs. 3(b) and 3(d)] x-ray dif-
fraction patterns. The I
NOYAG
data have been scaled to
match the number of shots in the I
YAG
case. The anisotropy
mainly originates in the scattering interference of the two
(heavy) iodine atoms. Parts of the zeroth order scattering
maximum and the first minimum (along the alignment
(a) (b)
(c) (d)
NoYAG YAG
E
YA G
E
TSL
E
YA G
E
FEL
Detector
p (arb. units)
x
p (arb. units)
z
-1
-1 1
1
0
0
p (arb. units)
z
-1
10
p (arb. units)
x
-1 10
Detector
FIG. 2 (color online). I
þ
ion images recorded with the ion-VMI
detector when (a), (b) the TSL or (c), (d) the LCLS ionize and
Coulomb explode the molecules. In (a) and (c), cylindrically
symmetric distributions from isotropic ensembles are observed
(the images are slightly distorted due to varying detector
efficiencies). In (b) and (d), the horizontal alignment of the
molecules, induced by the YAG, is clearly visible. In all
measurements, the YAG and the LCLS are linearly polarized
horizontally, parallel to the detector plane, and the TSL is linearly
polarized perpendicular to the VMI dete ctor plane.
photons/pixel photons/pixel
0.0
0.02
0.01
-0.01
-0.02
Simulation Experiment
(a)
(d)(c)
(b)
-150 -120 -90 -60 -30 -150 -120 -90 -60 -30
0.03
-0.03
0.02
0.0
-0.02
0.01
-0.01
-0.03
x
y
( )
O
( )
O
FIG. 3 (color online). Diffraction-difference I
YAG
− I
NOYAG
of
x-ray scattering in simulated (a) and experimental (b) x-ray-
diffraction patterns. Histograms of the corresponding angular
distributions on the bottom pnCCD are shown in (c) and (d),
respectively. Error bars correspond to 1σ statistical errors.
PRL 112, 083002 (2014)
PHYSICAL REVIEW LETTERS
week ending
28 FEBRUARY 2014
083002-3
direction α ¼ −60°) show up most prominently on the
bottom pnCCD panel. The simulated I
YAG
− I
NOYAG
image
has been calculated for a molecular beam density M of
DIBN molecules of M ¼ 0.8 × 10
8
cm
−3
. The error bars σ
correspond to the statistical errors from the I
YAG
− I
NOYAG
subtraction (σ ¼
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
I
YAG
þ I
NOYAG
p
). The histograms in
Figs. 3(c)–3(d) visualize the angular anisotropy which is
well beyond the statistical error in the experimentally
observed image [Fig. 3(d)], confirming the observation
of x-ray diffraction from strongly aligned samples
of DIBN.
To analyze which structural information can be derived
from the x-ray diffraction of isolated DIBN molecules, the
intensity IðsÞ in dependence of the scattering vector s ¼
sinðΘÞ=λ along the alignment direction α ¼ −60° is com-
pared to simulated models of different iodine-iodine dis-
tances. Θ is the scattering angle and 2Θ is the angle between
the beam direction and a giv en detector point [42]. Ab initio
calculations (
GAMESS
-US MP2/6-311 G** [43])predicta
value of 700 pm for the iodine-iodine distance. Figure 4
shows the experimentally obtained intensity profiles IðsÞ,
averaged over −70° ≤ α ≤−50°, together with simulated
IðsÞ profiles. Each curve is normalized to be independent of
the exact molecular beam density M of DIBN molecules,
which merely changes the contrast, i.e., the depth of the
minimum. Because of the relati v ely long wav elength
(620 pm) compared to the known iodine-iodine distance
(700 pm), the scattering extends to large angles and the first
scattering maximum from the iodine-iodine interference is
not covered by the detector in our setup. The experimentally
obtained IðsÞ is best fitted for an iodine-iodine distance of
800 pm. Figure 4 shows the simulated Ið sÞ for iodine-iodine
distances of 500, 700, 800, and 1000 pm. The inset of Fig. 4
depicts the calculated χ
2
values [44] in dependence of the
iodine-iodine distance. Because of the experimental param-
eters, as mentioned above, the scattering features are large
and vary only slightly within the recorded range of s values.
We note that the structural features of small molecules, like
DIBN, could be determined much more accurately with data
recorded at a shorter wavelength where the available s range
extends to cover sev eral maxima and minima. This would be
possible at wavelengths of 200–100pm,whichbecame
available at LCLS recently and will be available at upcoming
facilities, e.g., the European XFEL, in the near future.
We do not observe direct signs of radiation damage in the
diffraction data. While previous experiments aimed spe-
cifically at the investigation of x-ray induced damage in
strongly focused x-ray beams [8,45], here, we have actively
avoided that regime and performed the experiments using a
hundred times larger cross section of the x-ray beam. Under
these moderate-fluence conditions the damage can be
rationalized based on simple cross-section estimates for
photoionization and elastic scattering and is detailed in the
Supplemental Material [41]. Since the sample is replen-
ished for every XFEL pulse, the diffractive imaging signal
is only sensitive to the dynamics of damaged molecules
during the x-ray pulse (∼100 fs). Using a simple mechani-
cal model, we estimate that most (∼90%) of the diffraction
signal is due to (practically) intact molecules. A minor
fraction of the signal is due to damaged molecules with
small changes in molecular structure, which could not have
been resolved with the available x-ray wavelength. Damage
could even be mitigated using shorter (∼10 fs) duration
pulses; see Supplemental Material [41] for details.
Moreover, an appropriate trade-off between pulse duration,
pulse energy, and repetition rate would allow the recording
of atomically resolved x-ray diffraction patterns of mole-
cules within minutes [13]. At these high repetition rates,
one could directly observe femtosecond molecular dynam-
ics through snapshots for many time delays in pump-probe
experiments of electronic-ground-state chemical dynamics.
In summary, we demonstrate the preparation of strongly
aligned samples of polyatomic molecules at an XFEL
facility. We experimentally verify that the high-frequency,
far off-resonant x rays are an ideal probe of alignment of
molecular ensembles in a photoion momentum imaging
approach. The employed setup and conditions are appli-
cable for coherent diffractive imaging of single biomole-
cules or molecular ensembles. We show the possibility to
perform spatially resolved single x-ray photon counting.
Because of the weak scattering signal from small isolated
molecules, averaging of many shots is necessary and
possible for the observation of an analyzable diffraction
signal, on top of a large background from NIR photons. We
confirm that the angular structures in the single molecule
diffraction patterns were preserved during averaging and
FIG. 4 (color online). Comparison of experimentally obtained
intensity profiles IðsÞ along the alignment direction of the
diffraction-difference pattern I
YAG
− I
NOYAG
with simulated pro-
files. The experimentally obtained IðsÞ is best fitted (in terms of a
χ
2
test) with the model for an iodine-iodine distance of 800 pm
(inset: test-statistic χ
2
in dependence of the iodine-iodine
distance).
PRL 112, 083002 (2014)
PHYSICAL REVIEW LETTERS
week ending
28 FEBRUARY 2014
083002-4
that a diffraction pattern of isolated and strongly aligned
DIBN molecules was successfully measured beyond exper-
imental noise. Even with the experimentally limited range
of scattering vectors s, the heavy-atom distance derived
from the IðsÞ plot is in agreement with the computed
molecular structure, demonstrating the capability to extract
structural information for small molecules.
Our results provide direct evidence for the feasibility of
x-ray diffractive imaging of aligned gas-phase ensembles of
molecules. Analyzing radiation damage in detail shows that
damage effects in the diffraction pattern could be avoided
by using shorter x-ray pulses with lower fluences at higher
repetition rates. This would allow us to observe atomically
resolved snapshots of ultrafast chemical dynamics.
Combined with advanced molecular beam delivery tech-
niques, e.g., laser desorption or helium droplet beams,
considerably larger molecules could be delivered in cold
beams, isomer selected, and aligned, providing a bottom-up
approach toward the envisioned atomic-resolution single-
molecule diffraction experiments. In contrast to ultrafast
electron diffraction, pump-probe experiments with x-ray
pulses will not suffer from Coulomb-repulsion broadening
or pump-probe velocity mismatch and, hence, may permit
better time resolution, i.e., in the range of 10–100 fs.
Parts of this research were carried out at the Linac
Coherent Light Source (LCLS) at the SLAC National
Accelerator Laboratory. LCLS is an Office of Science
User Facility operated for the U. S. Department of Energy
Office of Science by Stanford University. We acknowledge
the Max Planck Society for funding the development and
operation of the CAMP instrument within the ASG at
CFEL. H. S. acknowledges support from the Carlsberg
Foundation. C. C. and P. J. acknowledge support from the
Swedish Research Council and the Swedish Foundation for
Strategic Research. J.C.H.S. and H.N.C. acknowledge NSF
STC Grant No. 1231306. A. Ru. acknowledges support
from the Chemical Sciences, Geosciences, and Biosciences
Division, Office of Basic Energy Sciences, Office
of Science, U.S. Department of Energy. D. R. acknowl-
edges support from the Helmholtz Gemeinschaft through
the Young Investigator Program. This work has been
supported by the excellence cluster “The Hamburg
Center for Ultrafast Imaging–Structure, Dynamics and
Control of Matter at the Atomic Scale” of the Deutsche
Forschungsgemeinschaft.
*
jochen.kuepper@cfel.de; http://desy.cfel.de/cid/cmi
a
Present address: Bruker AXS GmbH, Karlsruhe, Germany
b
Present address: ARC Centre of Excellence for Coherent
X-ray Science, School of Physics, The University of
Melbourne, Australia
c
Present address: Deutsches Elektronen-Synchrotron
(DESY), 22607 Hamburg, Germany
d
Present address: Physikalisch-Technische Bundesanstalt,
Bundesallee 100, 38116 Braunschweig, Germany
e
Present address: European X-Ray Free Electron Laser
(XFEL) GmbH, 22761 Hamburg, Germany
f
Stanford PULSE Institute, SLAC National Accelerator
Laboratory, 2575 Sand Hill Road, Menlo Park, California
94025, USA
g
Present address: PNSensor GmbH, 81739 Munich,
Germany
h
Present address: Institute for Molecules and Materials,
Radboud University Nijmegen, Heijendaalseweg 135,
6525 AJ Nijmegen, Netherlands
[1] R. Neutze, R. Wouts, D. van der Spoel, E. Weckert, and J.
Hajdu, Nature (London) 406, 752 (2000).
[2] J. Andruszkow et al., Phys. Rev. Lett. 85, 3825 (2000).
[3] H. N. Chapman et al., Nat. Phys. 2, 839 (2006).
[4] A. Barty et al., Nat. Photonics 2, 415 (2008) .
[5] Y. H. Jiang et al., Phys. Rev. Lett. 105, 263002 (2010) .
[6] P. Emma et al., Nat. Photonics 4, 641 (2010).
[7] C. Bostedt et al., J. Phys. B 46 , 164003 (2013).
[8] L. Young et al., Nature (London) 466, 56 (2010).
[9] H. N. Chapman et al., Nature (London) 470, 73 (2011).
[10] M. Marvin Seibert et al., Nature (London) 470, 78 (2011).
[11] J. M. Glownia et al., Opt. Express 18 , 17620 (2010).
[12] R. Boll et al., Phys. Rev. A 88, 061402 (2013).
[13] A. Barty, J. Küpper, and H. N. Chapman, Annu. Rev. Phys.
Chem. 64, 415 (2013).
[14] N. D. Loh et al., Nature (London) 486, 513 (2012).
[15] T. Gorkhover et al., Phys. Rev. Lett.
108, 245005 (2012).
[16] A. Barty et al., Nat. Photonics 6, 35 (2012) .
[17] F. Filsinger, G. Meijer, H. Stapelfeldt, H. Chapman, and J.
Küpper, Phys. Chem. Chem. Phys. 13, 2076 (2011).
[18] G. von Helden, T. Wyttenbach, and M. T. Bowers, Science
267, 1483 (1995).
[19] S. Trippel, Y.-P. Chang, S. Stern, T. Mullins, L. Holmegaard,
and J. Küpper Phys. Rev. A 86, 033202 (2012).
[20] F. Filsinger, J. Küpper, G. Meijer, J. L. Hansen, J. Maurer, J.
H. Nielsen, L. Holmegaard, and H. Stapelfeldt, Angew.
Chem. Int. Ed. 48, 6900 (2009).
[21] T. Kierspel, D. A. Horke, Y.-P. Chang, and J. Küpper, Chem.
Phys. Lett. 591, 130 (2014).
[22] F. Filsinger, U. Erlekam, G. von Helden, J. Küpper, and G.
Meijer, Phys. Rev. Lett. 100, 133003 (2008).
[23] H. Stapelfeldt and T. Seideman, Rev. Mod. Phys. 75, 543
(2003).
[24] J. C. H. Spence and R. B. Doak, Phys. Rev. Lett. 92, 198102
(2004).
[25] J. C. H. Spence, K. Schmidt, J. S. Wu, G. Hembree, U.
Weierstall, R. B. Doak, and P. Fromme, Acta Crystallogr.
Sect. A 61, 237 (2005).
[26] E. R. Peterson et al., Appl. Phys. Lett. 92, 094106 (2008).
[27] S. Pabst, P. J. Ho, and R. Santra, Phys. Rev. A 81, 043425
(2010).
[28] C. J. Hensley, J. Yang, and M. Centurion, Phys. Rev. Lett.
109, 133202 (2012).
[29] U. Lorenz, N. M. Kabachnik, E. Weckert, and I. A.
Vartanyants, Phys. Rev. E 86, 051911 (2012).
[30] B. Ziaja, H. N. Chapman, R. Fäustlin, S. Hau-Riege, Z.
Jurek, A. V. Martin, S. Toleikis, F. Wang, E. Weckert, and R.
Santra, New J. Phys. 14, 115015 (2012).
PRL 112, 083002 (2014)
PHYSICAL REVIEW LETTERS
week ending
28 FEBRUARY 2014
083002-5
