Charge Resonance Enhanced Ionization

• CREI is a laser-driven process in diatomic molecules where ionization is markedly enhanced at critical internuclear distances due to field-induced barrier suppression and electron localization.
• Experimental pump-probe studies of H₂⁺ reveal a double-peak structure in ionization rates, with peaks near internuclear separations of approximately 6.5 a.u. and 12 a.u., as evidenced by proton kinetic energy measurements.
• Theoretical analyses using fixed-nuclei models, wavepacket dynamics, and exact factorization show that non-adiabatic electron-nuclear coupling is crucial for accurately capturing the time-resolved ionization dynamics in CREI.

Searching arXiv for the specified CREI papers and closely related work to ground the article in the cited literature. Charge-resonance enhanced ionization (CREI) is a strong-field molecular-ionization phenomenon in which a diatomic molecule becomes markedly easier to ionize when its internuclear separation reaches a critical range. In the canonical case of $\mathrm{H}_2^+$, the effect arises because an intense laser field reshapes the molecular potential so that electron removal becomes highly efficient at particular values of $R$, rather than varying monotonically with bond length. The term is also used more broadly for related resonance-enhanced charge-transfer processes, but its standard meaning is the $R$-dependent enhancement of ionization in a laser-driven molecular ion, especially the long-studied case of $\mathrm{H}_2^+$ (Xu et al., 2015).

1. Definition and physical picture

CREI refers specifically to the strong increase of ionization probability in a diatomic molecule when the nuclei separate to a “critical” distance under an intense laser field. In $\mathrm{H}_2^+$, the physical picture is that the laser-dressed system resembles a tilted double-well potential: the electron can localize in one well, while the Coulomb field of the other proton lowers and distorts the barrier that confines it. Ionization becomes especially efficient when the molecule is stretched enough for significant localization, but not so stretched that internuclear coupling becomes too weak to shape the escape pathway effectively (Xu et al., 2015).

As $R$ increases from the equilibrium separation, barrier suppression improves and the ionization rate rises; beyond a certain range, the interplay of localization and charge-resonant coupling no longer favors the same tunneling pathway, and the rate falls again. This produces a critical region of $R$ with maximal ionization probability. Fixed-nuclei theories for $\mathrm{H}_2^+$ predict not merely a single maximum, but a double-peak structure in the $R$-dependent ionization rate, with maxima at about $R \approx 6.5$ a.u. and $R$0 a.u. in the experimental study that first resolved the structure (Xu et al., 2015).

A related formulation emphasizes that, in a strong field, the internal barrier between the two wells of $R$1 is lowered at certain critical internuclear separations, and that the field dresses the system into a pair of charge-resonant states. In that description, CREI is an $R$2-dependent enhancement of ionization from the cation caused by field-induced barrier suppression and charge resonance, rather than a generic multiphoton resonance (Jing et al., 2016).

2. Experimental demonstration in $R$3

The decisive experimental result was the observation of the long-predicted double-peak structure in the $R$4-dependent strong-field ionization rate of $R$5 using a few-cycle pump-probe experiment (Xu et al., 2015). In that experiment, a 6 fs, 750 nm laser pulse was split by a Mach-Zehnder interferometer into a strong pump and a much weaker probe. The pump had intensity $R$6, while the probe was about an order of magnitude weaker at $R$7. The pump ionized neutral $R$8, launching a nuclear wavepacket on the $R$9 potential curve; the probe then ionized the expanding molecular ion after a controlled delay.

The observable was the proton kinetic energy release (KER) following Coulomb explosion after double ionization. The experiment used the mapping

$R$0

with $R$1 taken as the measured bond-softening energy of $R$2 eV, so that

$R$3

This allowed conversion of measured KER spectra into effective $R$4-dependent spectra (Xu et al., 2015).

For parallel pump and probe polarizations, delay-dependent KER maps showed the enhanced-ionization channel only after the pump-created wavepacket had evolved for several femtoseconds. Around zero delay, the EI signal was absent, indicating that the pulses were short enough and their pedestals weak enough that enhanced ionization was not produced by pump or probe alone. The energy-integrated yield versus delay developed two clear peaks near 15 fs and 23 fs, while the corresponding KER branches contained one component whose central KER decreased with delay, tracking the expanding nuclear wavepacket, and another weaker, nearly delay-independent branch centered around 4.5 eV. The inferred $R$5-space maxima appeared at about $R$6 a.u. and $R$7 a.u., corresponding to KER peaks near 4.9 eV and 3 eV, respectively (Xu et al., 2015).

A critical control was the cross-polarized case. When the probe polarization was rotated by $R$8 relative to the pump, the energy-integrated yield no longer showed the double-peak structure; it rose monotonically and then saturated, and the enhanced-ionization mechanism was effectively absent. This demonstrated that the observed double peak in the parallel geometry was not a generic dissociation effect but a signature of CREI, which is highly anisotropic and requires the field to be oriented close to the molecular axis (Xu et al., 2015).

3. Theoretical descriptions: fixed nuclei, wavepacket dynamics, and exact factorization

The experimental results were supported by time-dependent Schrödinger equation simulations in a reduced-dimensionality model for the electron and nuclei,

$R$9

with a model Coulomb potential $\mathrm{H}_2^+$0 chosen to reproduce the molecular potential curves and the laser field represented through the vector potential $\mathrm{H}_2^+$1. A fixed-nuclei version of the model was also used to compute the $\mathrm{H}_2^+$2-dependent ionization rate directly, reproducing the same double-peak structure found experimentally (Xu et al., 2015).

This agreement established that the fixed-nuclei picture can be valid for the relevant observable in this regime, despite earlier suggestions that rapid nuclear motion would wash out the structure entirely. At the same time, the mismatch between calculated and observed relative peak heights suggested that more complete electron dynamics, including off-axis trajectories, may be required for a fully quantitative account of the second maximum (Xu et al., 2015).

A more stringent analysis of electron-nuclear coupling was developed using exact factorization in a model one-dimensional $\mathrm{H}_2^+$3 system (Khosravi et al., 2015). In that framework, the full wavefunction is written as

$\mathrm{H}_2^+$4

with the conditional nuclear wavefunction satisfying

$\mathrm{H}_2^+$5

The resulting exact electronic Schrödinger equation is

$\mathrm{H}_2^+$6

where $\mathrm{H}_2^+$7 is the exact time-dependent potential energy surface for electrons, or $\mathrm{H}_2^+$8-TDPES (Khosravi et al., 2015).

The exact potential is decomposed as

$\mathrm{H}_2^+$9

The first term extends the usual quasistatic potential to a quantum nuclear wavepacket, while the latter three terms encode electron-nuclear dynamical correlation beyond the frozen-nuclei picture. The traditional quasistatic approximation,

$\mathrm{H}_2^+$0

treats the nuclei as localized at $\mathrm{H}_2^+$1 and neglects wavepacket width, splitting, and non-adiabatic correlation (Khosravi et al., 2015).

The exact-factorization analysis showed that these omitted terms are not perturbative details. During ionization and dissociation, the exact $\mathrm{H}_2^+$2-TDPES can develop structures absent in $\mathrm{H}_2^+$3: shallower central wells, smaller and narrower outer barriers, and even double-well and later four-well structures as the nuclear wavepacket splits. The gauge-dependent term $\mathrm{H}_2^+$4 can introduce large steps that lower the barrier and strongly enhance ionization; $\mathrm{H}_2^+$5 can create barrier-like structures that modify tunneling; and $\mathrm{H}_2^+$6 can be confining and reduce ionization. The central conclusion was that non-adiabatic electron-nuclear coupling is crucial for accurate time-resolved ionization dynamics and ionization yields, so CREI cannot in general be understood correctly using a purely frozen-nuclei or quasistatic picture (Khosravi et al., 2015).

4. Time-resolved diagnostics and KER signatures

Because the second ionization maps the internuclear distance into proton kinetic energy, KER spectroscopy is a central diagnostic of CREI. In the $\mathrm{H}_2^+$7 pump-probe experiment, the conversion from $\mathrm{H}_2^+$8 to $\mathrm{H}_2^+$9 via the KER-to-$R$0 relation enabled direct comparison between measured delay-dependent spectra and the simulated expectation value $R$1, with good agreement (Xu et al., 2015).

In a broader study of laser-induced dissociative ionization of $R$2, CREI was identified as one of several mechanisms shaping proton KER spectra across wavelengths from 800 nm to 6400 nm (Jing et al., 2016). The relevant pathway was

$R$3

with nuclear motion on the $R$4 and $R$5 Born–Oppenheimer curves followed by Coulomb explosion on the $R$6 curve.

That study used field-ionization rates of $R$7 and noted that, at the field strength considered, CREI occurs around

$R$8

corresponding to Coulomb repulsion energies

$R$9

These energies were used as characteristic KER scales for identifying CREI contributions in the spectra (Jing et al., 2016).

The same work stressed that KER peaks do not originate from a single mechanism. It separated ionization following resonant dipole transitions between $R$0 and $R$1, CREI at critical separations, and ionization in the large-$R$2 dissociative limit. Many peaks were interpreted as composite structures obeying

$R$3

where $R$4 is the internuclear distance at the second ionization (Jing et al., 2016).

The Monte Carlo Wave Packet (MCWP) approach enabled trajectory-level separation of these mechanisms. The KER from a given trajectory was defined by projection onto Coulomb waves,

$R$5

and the full spectrum by summing over weighted trajectories,

$R$6

Within this formulation, CREI-related features were associated with second ionization near $R$7 or $R$8, while very low-energy peaks at long pulse durations and long wavelengths were assigned to ionization beyond the CREI region, at $R$9 (Jing et al., 2016).

5. Orientation, anisotropy, and dynamical nuclear effects

Anisotropy is central to the empirical identification of CREI. In the $\mathrm{H}_2^+$0 pump-probe measurements, CREI was strongly expressed only when the molecular axis was aligned close to the laser field direction; rotating the probe polarization by $\mathrm{H}_2^+$1 suppressed the mechanism, leaving only a monotonic delay dependence attributable to the decreasing ionization potential as the molecule dissociated (Xu et al., 2015). This establishes that enhanced ionization in $\mathrm{H}_2^+$2 is not isotropic barrier lowering but a geometry-sensitive process tied to the field orientation relative to the molecular axis.

Nuclear motion is equally important. The pump-probe experiment exploited the fact that the pump-created nuclear wavepacket takes about 15 fs to reach the outer turning point, making few-cycle pulses essential if one is to sample distinct internuclear separations before the CREI structure is blurred by averaging (Xu et al., 2015). Earlier suggestions that fast nuclear motion would make the double-peak structure impossible to observe were therefore directly addressed by pulse-duration control.

The exact-factorization analysis sharpened this point by showing that one must distinguish between the usefulness of the fixed-nuclei picture for identifying preferred $\mathrm{H}_2^+$3-values and the inadequacy of a purely quasistatic potential for full time-resolved dynamics. In that study, most ionization occurred when $\mathrm{H}_2^+$4 was roughly between $\mathrm{H}_2^+$5 and $\mathrm{H}_2^+$6 a.u., but an $\mathrm{H}_2^+$7-resolved analysis showed a strong ionization peak around $\mathrm{H}_2^+$8–$\mathrm{H}_2^+$9 a.u. depending on the pulse. This discrepancy illustrates that averages over the nuclear wavepacket can obscure which part of the split distribution actually ionizes (Khosravi et al., 2015).

To address this, the authors introduced the time-resolved, $R$0-resolved ionization probability

$R$1

with integration over the ionization region in $R$2. This quantity showed a pronounced peak in the same critical $R$3-range predicted by quasistatic CREI, but now in a fully dynamical description. A plausible implication is that CREI is best regarded not as a property of $R$4, but of the portions of the evolving nuclear distribution that enter the ionization-relevant configuration space (Khosravi et al., 2015).

6. Extensions, analogies, and conceptual scope

The concept of charge-resonance-enhanced ionization has also been invoked in systems beyond stretched one-electron diatomics, although the exact meaning varies with context. One extension concerns resonant charge transfer of hydrogen Rydberg atoms incident at a Cu(100) projected band-gap surface (Gibbard et al., 2015). There, charge transfer is strongly enhanced when the Rydberg electron energy matches a discrete surface image state. The mechanism is explicitly described as closely analogous to CREI: in both cases, charge transfer is greatly increased when two charge-localized states become energetically resonant, creating an efficient channel for electron transfer at larger separation than usual. The crucial distinction is that, in the surface problem, the relevant resonance is between an atomic Rydberg state and a discrete image-charge state within the projected band gap, rather than between two molecular centers in a laser-distorted double-well potential (Gibbard et al., 2015).

Another extension appears in a study of laser-driven ground-state $R$5, where the authors discuss mechanisms analogous to CREI in a multielectron, non-stretched, non-dissociating molecule at equilibrium geometry (Bauerle et al., 14 Aug 2025). In that work, intense ultrafast driving resonantly mixes the $R$6 and $R$7 orbitals, producing Rabi flopping, dynamic electron localization, oscillatory ionization rates, suppressed-ionization intervals, and multiple ionization bursts. The Rabi frequency is written as

$R$8

and the field-driven resonance is supported by Mollow sidebands in the HHG spectrum at offsets corresponding to $R$9 (Bauerle et al., 14 Aug 2025).

The same study characterizes ionization via the outgoing electronic flux,

$R \approx 6.5$0

and local binding via the time-dependent average local ionization energy,

$R \approx 6.5$1

It further analyzes localization through a time-dependent electron localization function based on

$R \approx 6.5$2

with

$R \approx 6.5$3

This suggests a broader conceptual reading of CREI as a field-driven resonance/localization phenomenon in which transiently localized charge redistribution modulates ionization, although the study itself explicitly frames the behavior as CREI-like rather than as the canonical stretched-$R \approx 6.5$4 mechanism (Bauerle et al., 14 Aug 2025).

7. Significance and open interpretive issues

The first direct observation of the double maximum in the $R \approx 6.5$5-dependent ionization rate of $R \approx 6.5$6 resolved a long-standing discrepancy between theory and experiment and validated the use of the fixed-nuclei picture for that observable in the relevant few-cycle regime (Xu et al., 2015). It also established strong-field molecular ionization as a probe of ultrafast nuclear motion, since the delay-dependent KER maps directly track the expanding nuclear wavepacket.

At the same time, subsequent theory showed that accurate time-resolved simulation of CREI requires more than identifying critical internuclear separations. Exact factorization demonstrated that the exact electronic potential can differ substantially from the traditional quasistatic potential because of non-adiabatic electron-nuclear coupling, and that these differences are crucial for predicting the onset time of ionization, the shape of the ionization burst, the saturation behavior of the yield, and the final ionization probability (Khosravi et al., 2015). A central interpretive tension therefore remains: fixed-nuclei or quasistatic analyses can correctly identify preferred $R \approx 6.5$7-ranges, yet can fail quantitatively or even qualitatively in the full dynamical problem.

A further issue is the interpretation of multiple low-energy KER features in dissociative double ionization. The MCWP analysis of $R \approx 6.5$8 emphasized that CREI must be distinguished from direct resonance-region ionization and from late ionization in the large-$R \approx 6.5$9 dissociative limit. Very low-energy peaks at long wavelength and long pulse duration are not CREI peaks in the strict sense, even though they also arise from strong-field ionization of an expanding molecular ion (Jing et al., 2016). This distinction is essential for assigning KER structures correctly.

In its narrow sense, CREI remains the $R$00-dependent enhancement of ionization in a laser-driven diatomic cation, most rigorously established for $R$01. In a broader sense, the term and its analogies motivate a class of resonance-assisted charge-transfer and localization phenomena in which external fields, molecular geometry, and correlated electron-nuclear dynamics jointly create preferred ionization channels. The literature indicates both the durability of the original $R$02 paradigm and the need for care when generalizing it to multielectron, surface-mediated, or non-dissociative settings (Xu et al., 2015).