Atomic excitation and ionization in strong laser fields

Recent experiments on short-pulse (duration of 100 fs), high-intensity (peak intensity of 10^13 Watts/cm^2) laser irradiation of noble gases have found significant amounts of ionization and residual excited-state populations to occur simultaneously. These observations have caused some controversy because the prevailing "shell" model asserts that ionization under such conditions is dominated by multiphoton ionization mediated by an excited state that is field-shifted into resonance at the peak of the pulse. In this model, ionization is viewed as a coherent process, in which the resonant intermediate state greatly enhances multiphoton ionization of the initial state without itself becoming populated. It is difficult to reconcile efficient ionization with the survival of this intermediate state after the laser pulse has passed.

We have investigated this effect in a model system by integration of the time-dependent Schroedinger equation.

This figure shows the energy levels of the model atom along with the size of the laser photon. Model atom energy levels are designated by labels E1, E2, etc., where E1 is the ground state. Each level has definite parity with the ground state having even parity and each succeeding level has a parity opposite that of its predecessor. On the right one sees the shifted energy levels of this atom in the laser field, as a function of the field intensity. Note the six-photon resonance condition between the shifted ground state and the level E5, just below 2 x 10^13 Watts/cm^2. This resonance is responsible for most of the ionization observed in the experiments.

This figure shows the population of level E5 after the passage of the pulse as a function of peak pulse intensity. Dashed line: numerical calculation; solid line: Landau-Zener-type treatment of two-level system. Note that the population rises by four orders-of-magnitude at the resonance intensity after which it executes small oscillations around a slowly increasing value. These oscillations are due to interference of the amplitudes for resonant excitation on the rising and falling edges of the pulse.

Main conclusions

Ionization occurs when the resonance condition obtains, and so the total ionization yield should be proportional to a time interval t during which the laser intensity is "close" in some sense to the resonance intensity Ir. For realistic pulse profiles, t is exhibits a sharp maximum when the peak intensity is comparable to Ir, and eventually decreases with increasing peak intensity. Thus the ionization rate eventually decreases with increasing intensity.

On the other hand, the probability for multiphoton excitation increases slowly with peak intensity. This means that for a given laser pulse, ions will come predominantly from a spatial "shell," whereas excited states will be produced more or less uniformly within the volume enclosed by that shell. Thus the ratio of net excitation to net ionization will be much larger than would be the case in a spatially homogeneous field. We believe this fact accounts for much of the apparent inconsistency of the data between experiments that measure excited-state populations and ions.

Go back