Atomic, Molecular and Optical Physics

Not Found

Fig. 1: Sketch of the resonant-Auger process in a laser-dressed atom. The optical field couples core-excited states and produces sidebands on the ejected electron energies.

The AMO Physics program seeks quantitative understanding of x-ray interactions with atoms and molecules from the weak-field limit explored at Argonne's Advanced Photon Source (APS) to the strong-field regime accessible at the Linac Coherent Light Source (LCLS) and other x-ray light sources around the world. We perform experimental and theoretical research on x-ray and inner-shell processes, laser control of atomic and molecular motions, laser-pump/x-ray-probe studies of molecular and chemical dynamics, and x-ray diffractive imaging of optically trapped nanoparticles. Theory is a key element of our program by predicting phenomena that motivate experiments and by simulating measured results. The AMO group is particularly active at APS beamline 7-ID where we use high-repetition-rate lasers for pump-probe measurements on the photodynamics of solvated molecules and x-ray/ion coincidence measurements on core-hole decay dynamics of gas-phase molecules. Topics of LCLS experiments include multiple-photon ionization, femtosecond pulse characterization, and x-ray-pump/x-ray-probe studies of molecular dynamics. Theoretical studies include laser-dressed atoms in weak and strong x-ray fields and intense x-ray interactions with atoms, clusters, and nanoparticles. Brief summaries of our current research projects are given below.

High intensity x-ray science

Multiple photon ionization

The AMO group performed some of the first experiments on intense x-ray interactions with atoms and molecules using the LCLS x-ray free-electron laser (XFEL). A dramatic early observation was the stripping of all ten electrons from atomic neon by single LCLS pulses. The process of sequential absorption of six photons and four Auger decays is illustrated in Fig. 2. The experiments also demonstrated efficient hollow-neon production by sequential photoionization of both K-shell electrons prior to Auger decay. In another experiment on atomic neon, a single x-ray pulse first ejected a valence electron and then excited a K-shell electron to the valence hole. This was the first example of using intense x rays to open and excite a ``hidden resonance,'' which is a technique we are exploiting in studies of molecular core-hole decay dynamics.

PAPA schematic

Fig. 2: Illustration of sequential photon absorption and Auger decay in atomic neon. The eight 2s and 2p electrons can be ejected using 800-eV photons, which is below the 870-eV K-edge. K-shell photoionization and Auger decay are the dominant processes at higher x-ray energies. The K-shell ionization energy and the core-hole lifetime both increase with charge state due to depletion of valence electrons. These factors affect the dependence of ion charge-state yields on x-ray energy and pulse duration.

The capabilities of the LCLS and other XFELs are continually improving with increased control of pulse durations, reduced bandwidths and increased coherence by seeding, and schemes to produce two or more pulses with variable delays and colors. Our recent experiments seek to exploit these properties for studies of multiple-photon ionization and core-hole decay dynamics.

LCLS pulse duration measurements

Not Found

Fig. 3: Demonstration of THz streaking of LCLS x-ray pulses: the kinetic energies of K-shell photoelectrons of neon generated with 1.1-keV photons are modulated by the THz streaking field. The technique can be used to measure the XFEL pulse durations.

The LCLS can produce very intense x-ray pulses believed to be as short as a few femtoseconds, but precise determination of the pulse durations is a research topic in itself. Any measuring scheme is rendered even more challenging by the operating mode of LCLS, called SASE (Self Amplified Spontaneous Emission), which makes it a purely chaotic source and ultimately requires a single shot measurement of every shot to get a full characterization of the source properties. In addition, the SASE operation produces an inherent temporal jitter between the x-ray pulses and any other laser source operating in parallel, which limits greatly the resolution of pump-probe techniques commonly used in time-resolved measurements.

The main route followed by our collaboration to get a handle on those temporal properties consists in transferring the time properties of the x-ray pulses to electron wave-packets produced during ionization or subsequent Auger decay of a gas target. The simultaneous presence of a strong laser field (operating in the visible, IR or THz region) modifies the energy spectrum of those electron wave-packets in a deterministic way, as shown in Fig. 3. The laser field "streaks" the electron energies as a function of the time of emission so that an electron energy measurement encodes time information.

Self-referencing time domain measurements of femtosecond inner shell dynamics

Not Found

Fig. 4: Concept for self-referenced timing measurement: inside the streaking ramp, the x-ray pulse emits photoelectrons whose energy is modified by the laser field, allowing referencing of the x-ray pulse inside the ramp. The electrons emitted by the process of interest (e.g. Auger decay) are also streaked and independently positioned relative to the ramp, and thus relative to this particular x-ray pulse. Statistics can then be accumulated to yield the time profile of the emission.

The availability of ultrashort x-ray laser pulses opens the potential of following ultrafast electronic and nuclear movements with atomic resolution. However, the timing jitter between the x-ray pulses and any external source can dominate the temporal resolution of the measurements. We propose a potentially very general technique to overcome this limitation when the observables are electrons (either photoelectrons or Auger electrons), by extending the laser streaking technique to measure simultaneously both the timing of direct photoionization from the x-ray pulse and the time dependent feature of interest. The concept is illustrated in Fig. 4.

Measuring the photoionization event allows one to create, on a shot by shot basis, an absolute reference for the timeline of events being triggered by that pulse. Because this timing is extracted for every shot, events with low statistics can then be correctly added up to yield high quality data in the time domain. The objective is to measure, for every shot, the streaked energy spectra of both the photoelectron and the other electrons of interest. A streaked photoelectron spectrum collected with good statistics allows determining, for a particular shot, the relative position inside the streaking ramp. Then, determining the amount of streaking experienced by the other electrons that are detected can directly indicate their emission time inside the ramp and relative to the x-ray pulse.

Hetero-site-specific femtosecond-time-resolved dynamics

Not Found

Fig. 5: Sketch of the dynamics in XeF2 triggered by a core-hole created in Xe. We show an example of how we can track a specific intermediate path by using the probe pulse to resonantly excite F to the induced holes.

Core-hole decay in molecules is a fundamental process far less understood than in isolated atoms. When an x-ray photon is absorbed by a molecule, a core-shell electron can be promoted to an unoccupied valence orbital or to the continuum, leaving a core-hole in the molecule. The core-hole decay will trigger a cascade of processes all occurring at the same time: Auger electron emission, fluorescence photon emission, dissociation/fragmentation of the ion/neutral molecule, energy transfer, and charge transfer. This complex process occurs on the femtosecond time scale. We have developed an experimental approach to follow the intermediate hole-state dynamics by using the capability at the LCLS to generate two femtosecond x-ray pulses of different photon energies with variable time delay. Figure 5 illustrates the two-pulse method for the case of XeF2, a molecule we have studied using x-ray/ion coincidence spectroscopy at the APS. The first x-ray pulse initiates a vacancy cascade by ejecting a Xe 3d electron. After a controlled delay, the second pulse resonantly excites a F 1s electron, i.e., to a hidden resonance. Corresponding theoretical work is in progress to simulate the core-hole excitation and decay transitions.

Radiation damage in nanoparticles by intense x-ray interactions

Not Found

Fig. 6: Processes in x-ray-nanoparticle interaction included in MC/MD model. ES: elastic x-ray scattering, P: photoionization, L: Lattice dynamics, EI: electron-impact ionization, A: Auger (Coster-Kronig) decay and F: fluorescence.

Intense, femtosecond XFEL pulses have been shown to be useful for serial crystallography, imaging biomolecular crystals of sub-micron dimensions. However, radiation damage induced by high intensity x-ray radiation in these crystals is unavoidable. In particular, x-ray photoionization, inner-shell transitions (Auger and fluorescence) and secondary ionization can take place with high probability within the pulse duration, and these processes can damage the structure of the molecules and potentially limit the usefulness of XFELs for molecular structural determination. We devised a theoretical model to investigate the fundamental mechanisms of the damage in clusters of varying size, with an ultimate aim of understanding phenomena from molecular to the nanometer scale, perhaps eventually approaching 1 MDa in size.

In order to obtain an atomistic view of the dynamical x-ray damage processes and the subsequent structural distortion on the target system throughout the x-ray pulse, we use a combined Monte-Carlo/Molecular-dynamics (MC/MD) computational model (see Fig. 6). The rates of all inner-shell transitions and the cross sections of photoionization of each subshell are obtained with Hartree-Fock-Slater calculations. During the x-ray pulse, the occurrences of the photoionization, inner-shell decay processes, electron-impact ionization and the site of their occurrences are treated by Monte-Carlo type methods. Subsequently, the dynamics of the photoelectrons, Auger electrons, secondary electrons and the atoms/ions in the systems are tracked using molecular dynamics methods. The advantage of this MC/MD model is that it allows us to compute the time-dependent coherent x-ray diffraction pattern of the target system by tracking the position of the delocalized electrons and the configuration (electronic configuration and charge state) and positions of atoms/ions. In addition, we can monitor the fluorescence spectrum, charge-state distribution, energy spectra, lattice dynamics, and dynamics of photoelectrons, Auger and secondary electrons.

X-ray Probes of Molecular and Chemical Dynamics

Photoinduced metal-to-ligand charge transfer and spin crossover in aqueous [Fe(bipy)3]2+

Not Found

Fig. 7: X-ray absorption (XAS), x-ray emission (XES), and x-ray diffuse scattering (XDS) measurements were made on the laser-induced high-spin state of aqueous [Fe(bipy)3]2+. High precision data were recorded by operating the laser at 3.26 MHz, one-half of the 6.52 MHz x-ray pulse rate, so that every x-ray pulse was used for the laser-on and laser-off data sets. A comprehensive picture of the photodynamics emerged by combining the three x-ray techniques.

We have implemented a high repetition rate (54 kHz--6.5 MHz), high power (>10 W) laser system at APS beamline 7ID for laser-pump/x-ray-probe studies of photoinduced molecular dynamics. Our end station consists of a liquid jet of solvated molecules, laser transport and focusing optics, and Kirkpatrick-Baez x-ray mirrors to implement a time-resolved x-ray microprobe. Results are shown in Fig. 7 for laser-induced metal-to-ligand charge transfer and spin crossover in aqueous [Fe(bipy)3]2+. Determinations of the spin states, electronic states, and atomic structures of the ground state and laser-excited state were derived from x-ray absorption, x-ray emission, and x-ray diffuse (liquid) scattering measurements. Time-resolved measurements with high signal:noise are obtained by efficient use of the x-ray flux at high repetition rates. Experiments using picosecond x-ray pulses at APS complement experiments using femtosecond x rays at the LCLS and SACLA XFELs

Molecular response to x-ray absorption and vacancy cascades

Not Found

Fig. 8: Hit 1 vs. Hit 2 ion time-of-flight scatter plot following Xe 1s ionization of XeF2. The diagonal lines are from the two energetic F ions. The vertical lines are events in which a zero-kinetic-energy Xe ion is detected in Hit 2.

X-ray absorption by a heavy atom produces an inner-shell hole that decays in a multi-step process with the ejection of fluorescent x rays and Auger electrons that results in high charge states on the atomic ion. In molecules, the inner-shell hole and the holes produced by the first decay steps remain localized, but eventually charge is redistributed to neighboring atoms and the system Coulomb explodes. This process is responsible for x-ray damage in molecules and materials and has been exploited for targeted destruction of malignant cells. To study the physics of core-hole dynamics in small molecules, we use APS x rays to ionize K-shell electrons from a heavy atom such as Br, I, and Xe of the molecules IBr, CH2BrI, and XeF2. A fluorescent x-ray from the K-shell hole is detected and triggers an x-ray/ion coincidence measurement of the fragment ions. Figure 8 shows an example of a scatter plot of ion coincidences in XeF2. The data yield the ion charge state distributions and kinetic energies released to the ions, and comparisons are made with calculated core-hole decays. Femtosecond XFEL pulses are also being used to explore core-hole decay dynamics.

Laser Control of Atoms, Molecules, Nanoparticles, and X-ray Processes

Optical control of resonant Auger processes

Not Found

Fig. 9: Simulated excitation and resonant Auger decay of Ne+ transition 2p5 ↔ 1s-12p6 by a coherent 3π x-ray pulse that transfers the entire ground-state population to the excited state, drives it back to the ground state, and back up to the excited state The temporal profile of the excited-state population is shown in (a) along with the vector potential of an optical field that generates sidebands on the Auger electron energies. The resulting Auger electron sideband distributions are shown in (b). Asymmetries appear in the angle-resolved sideband distributions due to interference between different quantum paths taken by the Auger electron.

Optical lasers are extensively used to control the populations of low-energy atomic and molecular states, i.e., coherent control of valence electronic states and molecular vibrational and rotational motions. We are exploring the use of optical lasers to control x-ray and inner-shell processes. Despite the short lifetime of a K-shell hole in Ne (2.4 fs), we have demonstrated theoretically and experimentally that an intense, 800-nm laser field can couple core-excited states, e.g., transfer population from the 1s-13p to the 1s-13s state. In further theoretical work, we have demonstrated that this population transfer will be manifested in the resonant Auger electron spectrum (see Fig. 1). The simulations also show that the angular anisotropies of the Auger electrons will be imprinted on the sidebands that arise when continuum electrons interact with an optical field.

As XFEL sources mature and control increases over pulse properties, it is expected that highly-coherent x-ray pulses will allow population control of core-excited resonant states, as illustrated in Fig. 9(a). Our simulations demonstrate how this capability can be combined with optical lasers for coherent control of x-ray and inner-shell transitions. Asymmetries in sidebands will appear due to interference between different multiphoton quantum paths taken by Auger electrons (see Fig. 9(b)). The asymmetries are very sensitive to the x-ray and optical parameters due to the intrinsic coherence of the process. The effects can be applied to future pump-probe experiments, optical control schemes of x-ray absorption, and x-ray pulse characterization in XFELs.

X-ray diffraction imaging of optically trapped micro- and nano-particles

Not Found

Fig. 10: Sketch of a setup for x-ray diffraction of particles held in an optical trap. Figure taken from D. Cojoc et al., Appl. Phys. Lett. 97, 244101 (2010).

Coherent x-ray diffraction imaging (CXDI) is a sensitive method for imaging crystalline and noncrystalline materials at high spatial resolution. A major barrier to applying CXDI to freestanding micro- and nano-scale objects is their tendency to freely move within the intense beam of the APS. Typically such objects must be securely bonded to a substrate, which can alter their internal structures. A solution to this problem may be to hold the object in optical trap. A sketch of such a setup is shown in Fig. 10. Our optical trapping setup contains a built-in light detector that is able to measure the displacement of trapped particles with sub-nanometer accuracy. We have used it to record optical images of several types of nanoparticles. The trap will be installed at the CXDI beamline 34-ID-C. Our prototype sample will be zinc oxide microcrystals which have been extensively studied with CXDI and have also been stably trapped in our optical tweezers. Successful implementation of optical tweezers for sample alignment and orientation will enable an entirely new class of samples to be studied with CXDI including fully hydrated biological samples.