What can we learn from high harmonic generation about molecular electronic structure and dynamics? On our way to answering this question we explored new paths in attosecond research. Most important was our dicovery, that many electronic states, analogous to the ionization of multiple orbitals, participate in high harmonic generation. This notion is now generally accepted and can be used for attosecond preparation of molecular electronic wavepackets. Our recent results on water show that lower lying orbitals leave dynamic traces in the harmonic spectra of different isotopes. We have also shed new light on phase matching effects shaping harmonic spectra. Our studies show that absolute harmonic are influenced by the macroscopic propagation. We are further exploring electronic structures on asymmetric top molecules like SO2 and H2O. Our recent results indicate that high harmonic spectra are especially sensitive to subtle alignment changes and pave the way for future studies of excited state non-Born Oppenheimer dynamics on those systems.
Attosecond Source Development:
We are currently building a laser source and beam line to produce high intensity XUV pulses with sub-femtosecond pulse durations. A 500 as pulse has a minimum coherent spectral bandwidth of ~3.5 eV, which can only be achieved with higher frequency electromagnetic radiation, e.g. XUV light. Strong-field-driven high harmonic generation (HHG) has proven to be a robust method for generating IAPs. HHG from laser pulses containing multiple laser cycles produces attosecond bursts of VUV/XUV radiation every half cycle of the driving infrared laser pulse. In order to create an IAP, one needs to filter out the attosecond burst from a single cycle of the driving laser pulse. There are various methods to accomplish this goal, which broadly fall into one of two categories, gating and spatio-temporal coupling (STC) schemes. In a gating scheme the driving laser pulse is manipulated either in spectrum, polarization, or waveform to suppress XUV bursts from all but a single laser cycle. Using gating techniques, pulses as short as 70~as have been generated and characterized. Conversely, in a spatio-temporal coupled laser field, spatial properties of the beam vary in time; for example, the direction of propagation can vary throughout the laser pulse duration, which is referred to as ultrafast wavefront rotation (WFR). An intense pulse with this particular STC will produce XUV bursts (or beamlets) every half laser cycle, but the beamlets will propagate in different angular directions (much like the beam from a lighthouse).
For pump/probe applications, one requires enough flux in a single attosecond burst to drive a resonant two-photon process. To date, there have only been a handful of demonstrations of IAPs with this level of XUV flux. The most intense IAP to date was generated using a high energy, low repetition rate (10 Hz) Ti:Saph laser system. Conversely, HHG pulse trains with total energy in the hundreds of nanojoule and even microjoule range have been readily used, and exploring two-photon interactions using these sources is becoming common. We propose to build a source that uses WFR to angularly separate attosecond-bursts (beamlets) from different cycles of the laser pulse and then re-time the bursts using a specialized split-mirror. This should greatly enhance the overall IAP energy, and produce nearly 200 nJ of XUV energy, assuming 30 mJ of infrared pulse energy. We estimate that this method will produce bursts with a duration between 300-500 as. Perhaps with further engineering, we could use this method to push to the 100 as level.
Impulsive Stimulated Raman Scattering (creating coherent electronic excitations):
Impulsive stimulated Raman scattering (ISRS) with pulse durations of tens of femtoseconds has been used with gas phase targets to create coherent excitations of rotational and vibrational motion, i.e. rotational and vibrational wave packets. We are working to extend these techniques to drive coherent electronic redistribution via ISRS. Given the recent advances in IAP source and FEL development, pulses with 2-10 eV of coherent bandwidth are becoming more common, and these are precisely the source one needs to drive ISRS. Recently we have calculated the ISRS probability for valence electron redistribution in Na atoms using few femtosecond XUV pulses, and found that above ~1015 W/cm2, the Raman excitation probability becomes larger than the single photoionization probability, as one might expect from perturbation theory. We are also interested in using x-ray pulses from LCLS to drive electronic state redistribution.
Resonant Raman excitation, whether by HHG laser pulses or soft x-rays from LCLS-II, should be a means to study coherent electron dynamics in molecules.
Transient Impulsive Electronic Raman Redistribution
S. Miyabe and P. H. Bucksbaum
Phys. Rev. Lett. 114 143005 (2015) DOI: 10.1103/PhysRevLett.114.143005
Multi Orbital Harmonics
The high harmonics of water can be used to track sub-femtosecond nuclear motion launched via ionization of the inner valence 3a1 orbital. This introduces a new method to find multi-orbital contributions to high harmonics. We observed nuclear and electronic motion of the water molecule on the attosecond time scale by comparing high harmonic spectra of water (H2O) and heavy water (D2O) (see c in Figure). Both the highest energy occupied orbital (HOMO) and also lower orbitals can participate in HHG, as we showed on the example of the nitrogen molecule in 2008. In water, the more deeply bound HOMO-1 ionization launches a wave packet that straightens the bend angle (a and b in Figure). The ionization from the lone pair HOMO orbital does not launch a motion in the molecule. The efficiency of HHG emission from HOMO-1 is then governed by the spatial overlap of the ionic state nuclear wave packet and the neutral vibrational ground state following recombination. This decreases as the bent molecule straightens out. The loss of overlap is less pronounced for the slower moving heavier isotope, so the harmonic ratio of H2O and D2O maps the bond motion. This observation shows how HHG can record rapid motion, and also reinforces growing evidence that HHG from multiple orbitals is not unusual. Additionally, the method of isotope marking allowed us to infer the multi-orbital character of strong field ionization without using any rotational or vibrational laser preexcitation. Our initial report on lower orbital harmonics in nitrogen have triggered many reports on multi orbital harmonics. Adding the case of water, we now believe that the harmonics from lower orbitals are the rule.
Strong field ionization to multiple ionic states of water
J. P. Farrell, S. Petretti, J. Foerster, B. K. McFarland, L. S. Spector, Y. V. Vanne, P. Decleva, P. H. Bucksbaum, A. Saenz, and M. Gühr
Phys. Rev. Lett., 107, 083001 (2011)
Further reading on multi-orbital ionization in strong fields:
Getting molecular electrons into motion
Science, 335, 1314 (2012)
High Harmonic Generation from Multiple Orbitals in N2
B. K. McFarland, J. P. Farrell, P. H. Bucksbaum and M. Gühr
Science, 322, 1232 (2008)
High harmonic Bragg gratings
Transient gratings are generally applied in the IR to UV range to overcome the reduced sensitivity problem on excited states. Two excitation pulses, intersecting under a small angle, create an excitation grating in the sample while a third (probe) pulse is diffracted from the grating. The diffracted signal has high sensitivity to excited state dynamics. We implemented the grating scheme for HHG where the harmonics are not only deflected into a sideband, but also dispersed in angle within a diffraction order, to enable HHS analysis (see figure above). This is achieved by enlarging the angle to 180°, resulting in a shorter grating period d, which disperses the harmonics to distinguishable angles without an additional grating element. Thereby we achieve a Bragg grating that is highly selective in its wavelength acceptance resulting in the dispersed and distinguishable harmonics.
Strongly Dispersive Transient Bragg Grating for High Harmonics
J. P. Farrell, L. S. Spector, M. B. Gaarde, B. K. McFarland, P. H. Bucksbaum and M. Gühr
Optics Letters, 35, 2028 (2010)
Harmonics from aligned asymmetric tops
The ability to directly image the structure of the outermost electrons in molecules, and thereby view chemical reactions as they occur, is an important goal in molecular physics and chemistry. We study high-order harmonic generation (HHG) in impulsively aligned quantum asymmetric tops. We quantify the angular contributions of HHG emission, making use of the full rotational revival structure. We find a signal sensitive to all five prolate top revival types and to fractional and multiple revivals, providing a new view of polyatomic rotations. Our results show that not only the HOMO orbital shape, but also the orientation dependence of the recombination dipole controls the harmonic efficiency. This has implications for HHG-based tomographic imaging.
Quantified angular contributions for high harmonic emission of molecules in three dimensions
L. S. Spector, M. Artamonov, S. Miyabe, T. Martinez, T. Seideman, M. Guehr, and P. H. Bucksbaum
arXiv, arXiv:1207.2517v1 (2012)
How phase matching alters harmonic spectra - Argon Cooper minimum
The goal of high harmonic spectroscopy is to deduce information about electronic structure of target atoms or molecules from the shape and phase of a harmonic spectrum generated on that particular target. A general problem in this method is that the spectral information is not only containing the response of the single molecule/atom, but also the macroscopic sample response originating from the phase matching of harmonics. While phase matching is necessary to observe the harmonics, its possibly hazardous in the interpretation of harmonic spectra in terms of single molecule/atom response.
We have studied the influence of phase matching on spectral information using the Cooper minimum of argon as a spectral marker. We have developed a new spectrometer in our lab to observe the wavelength content and divergence of a harmonic spectrum (see figure). The positions of the harmonic target with respect to the laser focus (1.3 and 1.9 mm in the two graphs) lead to different phase matching conditions manifested in dramatically different spectral shape and divergence. The Cooper minimum is absent in the left panel, whereas its pronounced at 51 eV in the left panel (harmonic at 51 eV is less intense than the neighboring ones). We have collaborated with M. Gaarde and K. Schafer (both Louisiana State University) to find the origins for that behavior. We found that the interference of s and d channels in the recombination together with phase matching effects lead to different modulation depth and energy location or also the complete absence of the Cooper minimum structural feature. However, the spectral phase of the d-channel, also reflecting the Cooper minimum, is not altered by phase matching effects. The study cautions to interpret harmonic spectra straight forwards in terms of electronic structure. The best protection against artefacts from phase matching seems to compare two spectra with similar phase matching but different excitation conditions or isotope content of the target (as in our N2 or water studies).
Influence of Phase Matching on the Cooper Minimum in Ar High Harmonic Spectra
J. P. Farrell, L. S. Spector, B. K. McFarland, P. H. Bucksbaum, M. Gühr, M. Gaarde, K. Schafer
Phys. Rev. A, 83, 023420 (2011)