1
Recoil-Ion and Electron Momentum Spectroscopy:
Reaction-Microscopes
J. Ullrich
1
, R. Moshammer
1
, A. Dorn
1
, R. Dörner
2
, L. Ph. H. Schmidt
2
, H. Schmidt-Böcking
2
1
Max-Planck-Institut für Kernphysik, Saupfercheckweg 1, D-67119 Heidelberg, Germany
2
Institut für Kernphysik, Universität Frankfurt, August Euler Str. 6, D-60486, Germany
Abstract
Recoil-ion and electron momentum spectroscopy is a rapidly developing technique that allows
one to measure the vector momenta of several ions and electrons resulting from atomic or
molecular fragmentation. In a unique combination, large solid angles close to
π
4
and
superior momentum resolutions around a few percent of an atomic unit ( ..ua ) are typically
reached in state-of-the art machines, so-called Reaction Microscopes. Evolving from recoil-
ion and COLd Target Recoil-Ion Momentum Spectroscopy” (COLTRIMS), Reaction
Microccopes – the “bubble chambers of atomic physics” – mark the decisive step forward to
investigate many-particle quantum-dynamics occurring when atomic and molecular systems
or even surfaces and solids are exposed to time-dependent external electromagnetic fields.
The present review concentrates on just these latest technical developments and on at least
four new classes of fragmentation experiments that have emerged within about the last five
years. First, multi-dimensional images in momentum space brought unprecedented
information on the dynamics of single-photon induced fragmentation of fixed-in-space
molecules and on their structure. Second, a break-through in the investigation of high-
intensity short-pulse laser induced fragmentation of atoms and molecules has been achieved
by using Reaction Microccopes. Third, for electron and ion-impact, the investigation of two-
electron reactions has matured to a state such that first fully differential cross sections (FDCS)
are reported. Forth, comprehensive sets of FDCS for single ionisation of atoms by ion-impact,
the most basic atomic fragmentation reaction, brought new insight, a couple of surprises and
unexpected challenges to theory at keV to GeV collision energies. In addition, a brief
summary on the kinematics is provided at the beginning. Finally, the rich future potential of
the method is shortly envisaged.
2
1. Introduction
2. Kinematics: A Brief Summary
2.1 General Considerations
2.2 Fast Ion-Atom Collisions
2.3 Collisions with Photons
3. Imaging Techniques
3.1 Reaction Microscopes
3.1.1 Imaging of Ions
3.1.2 Imaging of Electrons
3.2 Target-Preparation
3.2.1 Supersonic Gas-Jets
3.2.2 Magneto-Optical-Traps (MOTRIMS)
3.3 New Developments
4. Results
4.1 Photons
4.1.1 Double Ionisation of Atoms: A Brief Summary
4.1.2 Photo Ionisation of Fixed-in-Space Molecules
4.2 Single and Multiple Ionisation in Intense Laser Fields
4.2.1 Single Ionisation: Recollision in the Tunnelling Regime
4.2.2 Double Ionisation: Non-Sequential and Sequential
4.2.3 Correlated Motion of the Electrons in the Sequential and Non-Sequential
Regimes
4.2.4 Final State Coulomb Repulsion
and Sub-Threshold Recollision
4.2.5 Transverse Momentum Exchange
4.3 Electron Impact Ionisation
4.3.1 Single and Double Ionisation
4.3.2 Ionisation plus Excitation
4.3.3 Laser Assisted (e,2e)
4.4 Ion Collisions
4.4.1 Electron Capture: Dynamics and Spectroscopy
4.4.2 Projectile Ionisation: A Novel Approach to (e,2e)-Experiments on Ions?
4.4.3 Ionisation by Slow Projectiles: Saddle Point Electrons
4.4.4 Ionisation by Fast Projectiles: Attosecond Pulses
4.4.4.1 Sinlge ionisation at small and large perturbation
4.4.4.2 Double ionisation at small and large perturbation
4.4.4.3 Multiple ionisation at large perturbation
4.4.5 A Short Summary
5. A View into the Future
3
1. Introduction
The present review tries to give an experimentally biased overview on the present state of
understanding and research of a tremendously fast developing field, namely the investigation
of the quantum-dynamics of fragmenting atoms and molecules. In striking contrast to the
profound theoretical knowledge based on very precise experimental data that has been
achieved in the investigation of the
static structure of atoms and molecules, even the most
simple and, thus, fundamental
dynamical problems still pose severe challenges to theory:
Only three years ago, it has been reported that single ionisation of the hydrogen atom by
electron impact, the most basic fragmentation reaction, has been solved in a mathematically
consistent way (Rescigno
et al (1999) and references therein, Bray (2002)). The methods
employed, a large scale partial wave expansion, made use of massively parallel
supercomputers. Until now however, it has not been demonstrated to be practicable for ion
encounters or electron impact at lower or even higher energies. More recently, three-
dimensional imaging of the electron emission for single ionisation of helium by fast bare ionic
projectiles brought to light severe discrepancies with existing theoretical descriptions in the
perturbative (Schulz
et al 2003) as well as in the non-perturbative regime (Moshammer et al
2001). At low collision velocities, where “saddle point electrons” were once predicted to be
emitted by Olson (1983, 1986), rich structures in the impact parameter dependent electron
momentum distributions (Dörner
et al 1996; Abdallah et al 1997, 1998; Edgü-Fry E et al
2002; Afaneh
et al 2002) are still not quantitatively explained by theory (Macek and
Ovchinnikov 1998; Sidky
et al 2000, Sidky and Lin 2001).
Until now, the more “complicated” complete disintegration of a helium atom, the simplest
many-electron system where correlation has to be taken into account (see e.g. McGuire
(1995,1997), Ford and Reading (1988, 1990), Bronk
et al (1998)), has been successfully
described theoretically on the level of FDCS only for fragmentation by single photons (for a
recent review see e.g. Briggs and Schmidt (2000)) or a fast electron impact (Kheifets
et al
1999, Dorn
et al 2001, 2002, 2002a). Helium double ionisation in the non-perturbative regime
induced by intense femtosecond laser fields or by strong, ion-generated attosecond pulses
seems to be far from being solved theoretically. Multiple ionisation finally, poses
insurmountable problems to quantum theory on the level of fully differential cross sections
and available data have to be compared to predictions of classical many-particle calculations
(Schulz
et al 2000).
4
In the recent past, essentially since less than a decade ago, the field was revolutionised from
the experimental point of view by the invention of advanced, innovative many-particle mo-
mentum imaging and projection techniques based on large area position- and time-sensitive
multihit electron and ion detectors. The integration of target preparation, projection tech-
niques and detector development (Martin
et al 1981, Sobottka and Williams 1988, Jagutzki et
al 1998) lead to todays Reaction Microccopes – the “bubble chambers of atomic and molecu-
lar physics” – developed by Moshammer
et al (1994,1996) and Ullrich et al (1995). They
enable to measure the vector momenta of several fragments (ions, electrons, molecular ions)
with unprecedented large solid angles, often reaching hundred percent of
π
4
, at extreme pre-
cision: Energy resolutions below
meV1 are achieved for slow electrons while ion momenta
are routinely recorded at the eV
µ
1 level, corresponding to a temperature of a few milli-
Kelvin (for the detection of low-energy electrons ( eV5< ) in coincidence with recoil-ions see
also Kravis et al (1996), Dörner et al (1996a,b) and Abdallah et al (1998)). Additional techni-
cal progress in the projectile beam preparation, namely the availability of nanosecond pulsed
electron or ion beams as well as intense pulsed photon beams from 3
rd
generation light
sources or kilohertz, ultra-fast strong laser systems, accelerated the data-taking efficiency de-
cisively. Now, not only “kinematically complete” measurements have become feasible but
moreover, fully differential cross sections can be projected out of huge data sets.
In parallel, despite of general problems, substantial progress has been achieved in the theo-
retical treatment of fragmenting Coulomb systems, driven by conceptual innovations as well
as by the dramatic growth of computational capabilities in recent years. For example, the ex-
terior complex scaling method mentioned above, even if not easily to be generalized, never-
theless did solve the fundamental three-particle Coulomb problem in excellent agreement with
experimental results. Moreover, convergent close coupling calculations as well as hyper-
spherical R-matrix methods combined with semi-classical outgoing waves are nowadays able
to reliably predict fully differential fragmentation patterns for photo double ionisation of he-
lium. Meanwhile, within the last three years, the close coupling technique has been success-
fully applied to describe double ionisation by charged particle impact at high velocities and
first successful attempts have been undertaken to implement higher-order contributions at
lower energies. In addition, S-matrix approaches to describe the interaction of strong laser
fields with atoms, numerical grid methods to directly integrate the Schrödinger equation, hid-
den crossing techniques for ion impact at low collision energies, time-dependent density func-
5
tional theory to approach “true” many-electron problems and many more were successfully
developed or applied in the recent past (see e.g. Ullrich and Shevelko (2003)).
Historically, Reaction Microccopes have emerged from “Recoil-Ion Momentum
Spectroscopy” (RIMS) and COLTRIMS, continuously developed since the first recoil-ion
momentum measurements by Ullrich and Schmidt-Böcking in Frankfurt (Ullrich 1987,
Ullrich and Schmidt-Böcking 1987, Ullrich et al 1988, 1988a). Few groups world wide, the
one at Kansas State University with Cocke and Ali, of Grandin and Cassimi at the GANIL in
Caen, the group at the University of Frankfurt with Schmidt-Böcking, Dörner, Mergel,
Schmidt, Jagutzki and others, those of Ullrich, Moshammer and Dorn at GSI, Freiburg
University and now at the Max-Planck-Institute in Heidelberg made decisive contributions
over 15 years to arrive at the present state of sophistication. The historical development as
well as the wealth of results obtained with RIMS, the earlier recoil-ion momentum
spectrometers and COLTRIMS, the high-resolution
π
4 -detection of the recoil ion alone
(sometimes COLTRIMS is also used as a synonym for simultaneous ion and electron
momentum spectroscopy) were summarized in detail in several previous reviews on the field
(Ullrich 1994, Ullrich et al 1997, Dörner et al 2000).
Therefore, and in the light of explosion-like progress within the last five years, the present
review exclusively reports on the most recent experimental results and technical develop-
ments which are not or rarely covered in the previous reviews. After a brief summary of the
kinematics in Chapter 2, the latest technical developments are described in Chapter 3. In
Chapter 4 single photon, intense laser, electron and fast ion impact induced fragmentation
processes are reported within four Sections. Compared to early work with single photons at
synchrotrons using Reaction Microccopes, research has been strongly evolving towards mo-
lecular physics, exploring the fragmentation dynamics of fixed-in-space molecules which is
described in Section 4.1. Since the last review we have witnessed the first successful recoil-
ion momentum measurement on intense laser induced break-up reactions of atoms performed
by Moshammer et al (2000) and Weber et al (2000). Due to the rapid development in laser
technology producing shorter and shorter pulses down to two optical cycles, achieving phase
stabilization within the pulse envelope, producing attosecond higher harmonic photon pulses
etc., this topic progresses extremely fast. Reaction Microccopes start to play a key-role in the
field and the whole Section 4.2 is devoted to it. Furthermore, electron impact induced two-
electron processes “just under way” as reported by Dörner et al (2000) have seen a break-
through since then with a set of kinematically complete (e,3e) or ionisation-plus-excitation
measurements and first successful attempts to investigate laser-assisted (e,2e) reactions, de-
