Angle-dependent interferences in electron emission accompanying stimulated Compton scattering from molecules

The study explores how stimulated Compton scattering (SCS) manifests in molecules at sub-keV photon energies, where dipole (A·P) and non-dipole (A2) interactions can be comparable. Prior kinematically complete experiments in helium revealed key differences between Compton scattering from bound versus free electrons, notably the dominance of the non-dipole A2 term at keV energies leading to suppressed forward photon scattering. At lower photon energies, the relative contribution of A2 decreases, suggesting possible interference between A·P and A2 pathways that could alter electron emission patterns. Molecules add orientation-dependent complexity to electron angular distributions. Leveraging the intensity and ultrashort duration of XFEL pulses, which enable nonlinear X-ray processes and multi-particle coincidence measurements, the authors target SCS in fixed-in-space H2 near ~500 eV to probe how interference between A·P and A2 affects electron emission directionality and its dependence on molecular orientation and scattering geometry.

Compton scattering from bound electrons exhibits much smaller cross sections than photoionization, posing experimental challenges, particularly for molecules. Recent helium experiments at keV energies used kinematically complete detection to characterize scattering, showing suppression of forward photon scattering due to A2 dominance. Theoretical and experimental advances with XFELs have enabled nonlinear X-ray phenomena, two-color schemes, attosecond pulses, and molecular-frame photoelectron imaging. Prior theoretical work on SCS and stimulated Raman processes in hydrogen atoms established the importance and treatment of non-dipole effects at high photon energies, and underscored the feasibility of using ultrashort, intense pulses. Molecular-frame photoelectron angular distributions (MFPADs) in other contexts have demonstrated strong orientation dependence, motivating investigation of SCS in molecules where interference between channels of different symmetry can be probed.

The study models SCS from fixed-in-space H2 in the fixed-nuclei approximation, justified for ultrashort pulses of 200 attoseconds where nuclear motion is negligible. The light–matter interaction includes both dipole and non-dipole terms, with the Hamiltonian H = He + A(r,t)·P + 1/2 A(r,t)^2. Two excitation scenarios are considered: (i) a single linearly polarized pulse (one color) where incoming and scattered photons propagate in the same direction, and (ii) two linearly polarized, counter-propagating pulses (two colors) allowing forward and backward scattering. The polarization is the same for both pulses, and a Gaussian temporal envelope is used. At ~500 eV photon energies (~15–20 a.u.), the dipole term A·P is treated within the dipole approximation, and the non-dipole term A^2 is retained to first order in (k1−k2)·r, keeping interaction terms up to order 1/c after an appropriate gauge transformation removes the A1+A2 term; this approximation is validated by prior studies showing negligible differences relative to higher-order treatments at even higher photon energies. The time-dependent Schrödinger equation is solved by a spectral method: the wavefunction is expanded in 6880 correlated two-electron states built from antisymmetrized products of H2 orbitals represented with radial B-spline functions and spherical harmonics. The angular momentum expansion is truncated at l = 16. The radial box extends to 60 a.u. with 280 B-splines. Typical configuration counts: ~280 for continuum states and 390–700 for bound states. Continuum states satisfy incoming scattering boundary conditions. Pulse parameters: single-pulse case uses central frequency ω1 = 20 a.u. (~544 eV), peak intensity 10^18 W/cm^2, duration 200 as; two-pulse case uses counter-propagating pulses with ω1 = 20 a.u. and ω2 = 15 a.u., same peak intensity and duration. Molecular-frame photoelectron angular distributions (MFPADs) are computed for specified geometrical configurations of molecular axis, polarization, and incidence directions; in some analyses, yields are integrated over electron energies from 0 to 1.0 a.u. near threshold to compare trends. Selection rules derived from D∞h symmetry indicate that A·P (two-photon) and A^2 transitions populate final continuum states of different inversion symmetry, enabling coherent interference when their amplitudes are comparable, which is central to predicting asymmetric angular distributions.

• SCS is significant near ~500 eV: Despite being below the ~2 keV threshold for maximum momentum transfer in standard Compton backscattering from an electron at rest, SCS is apparent due to the broadband attosecond pulses. Near threshold electron energies, SCS probabilities are about three orders of magnitude smaller than one-photon single ionization probabilities yielding ~20 a.u. electrons, and comparable to two-photon single ionization probabilities producing ~40 a.u. electrons; however, in the near-threshold region, contributions from one- and two-photon ionization are negligible compared to SCS. Unstimulated (spontaneous) Compton scattering is estimated to be roughly four orders of magnitude weaker than SCS under the same pulse conditions.
• Single-pulse geometry (same direction for incoming and scattered photons): Strongly asymmetric MFPADs arise from interference between A·P and A^2 terms and depend critically on molecular orientation. For a molecule parallel to the polarization or to the incidence direction, forward electron emission dominates; for a molecule perpendicular to both polarization and incidence directions, backward emission dominates; when perpendicular to polarization but parallel to incidence, forward emission is favored. Without non-dipole effects, MFPADs would be inversion-symmetric.
• Two counter-propagating pulses (backward photon scattering enabled): The ionization probability shows two peaks—a near-threshold peak (forward SCS) and a second peak at electron energy ≈ ω1 − ω2 − Ip ≈ 4 a.u., corresponding to backward SCS; pronounced oscillations appear up to ~7 a.u., especially in the A^2 contribution. The relative importance of A·P and A^2 reverses around ~4 a.u. MFPADs near ~4 a.u. reveal: (i) for molecule parallel to polarization, a slight preference for backward emission (contrast to single-pulse case); (ii) for molecule perpendicular to both polarization and incidence, no forward/backward preference (contrast to single-pulse backward dominance); (iii) for molecule perpendicular to polarization and parallel to incidence, electrons are preferentially ejected orthogonally to both directions and in the backward direction.
• Energy dependence of angular asymmetry: With polarization aligned to the molecular axis and two counter-propagating pulses, increasing electron energy drives a transition from forward-preferred to backward-preferred emission, reflecting a change in the relative phase between the interfering A·P and A^2 amplitudes. Overall, orientation, polarization, and pulse geometry can switch preferential electron emission among forward, backward, both, or orthogonal directions—behavior not seen in atomic targets—due to interference between dipole and non-dipole pathways.

The results directly address how interference between dipole (A·P) and non-dipole (A^2) interactions shapes electron emission in SCS from molecules at sub-keV energies. Because A·P and A^2 populate continuum states of different inversion symmetry, their coherent superposition produces pronounced asymmetries that are highly sensitive to molecular orientation and scattering geometry. This explains and predicts orientation-controlled switching of electron emission between forward, backward, and orthogonal directions. The findings contrast with atomic or free-electron Compton scattering, where forward emission trends are more uniform and non-dipole effects dominate only at higher energies. The identification of electron-energy-dependent phase shifts between channels further clarifies how emission directionality can invert with energy. Practically, the strong SCS signal relative to spontaneous Compton under XFEL conditions, combined with the ability to align molecules and tailor pulse geometries, suggests new routes to probe and manipulate ultrafast molecular electron dynamics via MFPADs at their natural timescales.

The study demonstrates that in SCS of ~500 eV photons from aligned H2, interference between dipole and non-dipole interactions generates large, orientation-dependent asymmetries in molecular-frame electron angular distributions. By tuning molecular orientation relative to polarization and incidence directions, electron emission can be directed forward, backward, both, or orthogonally. These effects, unique to molecules, are enhanced at few-hundred eV energies and are experimentally accessible with current ultrashort, intense XFEL pulses. Future work includes extending the light–matter interaction to all orders in 1/c to treat stronger non-dipole effects at higher photon energies, incorporating nuclear dynamics to handle longer pulses and vibrational resolution, and employing fully quantum X-ray/matter interaction models to evaluate and compare spontaneous Compton scattering at higher energies where it may compete with SCS.

• Fixed-nuclei approximation: Nuclear motion is neglected, valid for ~200 as pulses; for longer pulses (few femtoseconds), hydrogen nuclear dynamics can be significant and must be included.
• Truncated interaction: The non-dipole interaction is retained only to first order in (k1−k2)·r and terms up to O(1/c); higher-order terms are omitted, though prior studies suggest limited impact at the studied energies.
• Specific pulse parameters and geometries: Results are shown for select photon energies (ω1=20 a.u., ω2=15 a.u.), intensities (10^18 W/cm^2), durations (200 as), and geometrical configurations; generalization to other conditions requires further computation.
• Single molecular species: Only H2 is considered; extension to more complex molecules may reveal additional effects.
• Spontaneous Compton scattering is estimated rather than fully computed with a quantum electrodynamical treatment; a full comparison at higher energies remains for future work.