We introduce a unified and simplified theory of atomic double ionization. Our results show that at high laser intensities (I ≥ 10 14 watts/cm 2 ) purely classical correlation is strong enough to account for all of the main features observed in experiments to date.Short-pulse lasers with high peak intensities (10 14 ≤ I ≤ 10 16 , in watts/cm 2 ) now produce multiphoton generation of double ionization, the two-electron photoelectric effect, with surprising results. To summarize briefly, the experimental data show that two atomic (or molecular) outer-shell electrons are highly correlated when photoejected, with a double ionization rate that can be 1-million times higher than uncorrelated sequential theory [1] allows, so the process is called non-sequential double ionization (NSDI). The first laboratory results were reported in 1992 and 1993 [2,3], showing an anomalously high double ionization yield, the principal experimental signature of NSDI. Additional data is being reported from momentum spectroscopy experiments [4,5,6,7,8,9,10,11]. The momentum distribution data, along with the ion-yield data, serve as the benchmarks for various theoretical models.The mechanism that makes NSDI correlation so effective is far from settled, and theoretical exploration has been extensive [12,13,14,15,16,17]. Almost all existing calculations refer to, or are closely guided by, a single few-step rescattering model [18,19], which is based on an imagined picture in which one electron escapes the atom by quantum tunneling through a fieldlowered barrier and is then phase-coherently and classically forced by the laser away from and then back to the core where a quantum collision liberates both electrons at once (consistent with the term non-sequential). However, the patchwork of ad hoc elements typically employed [20] has not been claimed to make a complete, i.e., self-contained, theory. It is the purpose of this note to show that a self-contained theory exists that is compatible with essentially all prominent features of NSDI.Our theory is dynamically classical, and discards all aspects of quantum mechanics including tunneling. It is built on the need for strong electron correlation to explain NSDI, and so must be intrinsically a two-electron theory. We do not advocate such a theory for an electron that does not have the advantage of a strongly correlated partner. It turns out that entirely classical interactions are adequate to generate very strong two-electron correlation, as observed in NSDI, and quantum theory is not needed. Of course atoms are quantum objects but in such strong fields as are used for NSDI it is mainly electron physics rather than atomic physics that determines the experimental outcome. In this sense the early remark of Corkum [19] advocating the adoption of a plasma per- spective was quite appropriate.We note that a new form of energy analysis is very helpful. The graph in Fig. 1 shows the sum of kinetic energy, electron-nucleus binding energy, e-e correlation energy and laser field interaction energy for each of...

We use classical simulations to analyze the dynamics of nonsequential double-electron short-pulse photoionization. We utilize a microcanonical ensemble of 10(5) two-electron "trajectories," a number large enough to provide large subensembles and even sub-subensembles associated with double ionization. We focus on key events in the final doubly ionized subensemble and back-analyze the subensemble's history, revealing a classical slow-down scenario for nonsequential double ionization. We analyze the dynamics of these slow-down collisions and find that a good phase match between the motions of the electrons can lead to very effective energy transfer, followed by escape over a suppressed barrier.

We compare quantum mechanical and fully classical treatments of electron dynamics accompanying strong field double ionization. The major features seen in quantum mechanical simulations, including the double-ionization jets, are reproduced when using a classical ensemble of two-particle trajectories.

We present global potential energy surfaces for the three lowest triplet states in O(3P)+H2O(X1A1) collisions and present results of classical dynamics calculations on the O(3P)+H2O(X1A1)-->OH(X2pi)+OH(X2pi) reaction using these surfaces. The surfaces are spline-based fits of approximately 20,000 fixed geometry ab initio calculations at the complete-active-space self-consistent field+second-order perturbation theory (CASSCF+MP2) level with a O(4s3p2d1f)/H(3s2p) one electron basis set. Computed rate constants compare well to measurements in the 1000-2500 K range using these surfaces. We also compute the total, rovibrationally resolved, and differential angular cross sections at fixed collision velocities from near threshold at approximately 4 km s(-1) (16.9 kcal mol(-1) collision energy) to 11 km s(-1) (122.5 kcal mol(-1) collision energy), and we compare these computed cross sections to available space-based and laboratory data. A major finding of the present work is that above approximately 40 kcal mol(-1) collision energy rovibrationally excited OH(X2pi) products are a significant and perhaps dominant contributor to the observed 1-5 micro spectral emission from O(3P)+H2O(X1A1) collisions. Another important result is that OH(X2pi) products are formed in two distinct rovibrational distributions. The "active" OH products are formed with the reagent O atom, and their rovibrational distributions are extremely hot. The remaining "spectator" OH is relatively rovibrationally cold. For the active OH, rotational energy is dominant at all collision velocities, but the opposite holds for the spectator OH. Summed over both OH products, below approximately 50 kcal mol(-1) collision energy, vibration dominates the OH internal energy, and above approximately 50 kcal mol(-1) rotation is greater than vibrational energy. As the collision energy increases, energy is diverted from vibration to mostly translational energy. We note that the present fitted surfaces can also be used to investigate direct collisional excitation of H2O(X1A1) by O(3P) and also OH(X2pi)+OH(X2pi) collisions.