C-C. Hung, P.S. Krstic, and D.R. Schultz Controlled Fusion Atomic Data Center Oak Ridge National Laboratory Oak Ridge, Tennessee 37831-6373, USA

Owing to the use of argon as a diagnostic species in fusion plasma devices, we have computed the state-selective charge transfer cross sections (and other inelastic processes) for Ar^15-18+ colliding with atomic hydrogen at energies between 15 and 100 keV/u, using the classical trajectory Monte Carlo method. A tabular summary of the results is given here.

1. INTRODUCTION

The classical trajectory Monte Carlo (CTMC) technique has been used to compute the inelastic cross sections in collisions of fully stripped argon with atomic hydrogen. This approach is usually accurate to within about 25-100 percent for collision from velocities roughly equal to the orbital velocity of the target electron up to a few times this velocity. For capture to high lying states of the projectile (in the case of highly charged ion-impact), the lower limit of the range of applicability has empirically been found to extend to somewhat lower velocities. Certainly, alternative methods which are fully quantum mechanical would be very difficult to implement in this range of intermediate collision energy and capture to relatively high n-level.

To begin with, we have first checked to see that the current results are indeed in agreement with those published by Olson (PRA 24, 1726 (1981)) previously. Then, we have repeated the calculations at successively lower energies to reach the regime of primary interest in present plasma diagnostics.

Utilizing a model potential, we have also extended these calculations to other projectile charge states (i.e. Ar^15-17+). As long as no significant amount of capture occurs to very low n-levels of the projectile, the CTMC method should be reasonably applicable to these systems as well. That is, the distribution to which the target electron is captured shifts to lower n-levels of the projectile with the lowering of projectile charge state. In spite of using a model potential to simulate the proper screening of the projectile nuclear charge, the present CTMC approach can not take into account the detailed interaction of the target electron with any electrons carried by the projectile. When the CTMC code output shows significant capture to shells which are already populated on the projectile, the limit of applicability of the method is being reached. Populations obtained in higher n-levels may still be reasonable, but this must be confirmed through independent means. Of course, for the fully stripped ion, Ar¹⁸⁺, this effect is absent. For ions of only slightly lower charge state (e.g. Ar^15-17+), there is little or no capture to the ground (n=1) level and consequently data for these ions should be reliable to a similar accuracy. This criteria would limit how low in charge state the CTMC approach would produce reasonably reliable results.

2. RESULTS

Here we display in tables our results for 15-100 keV/u Ar^15-18+ + H giving the state-selective charge transfer cross section explicitly. Throughout, error bars are indicated by the statistical uncertainty (one standard deviation) in the Monte Carlo statistics. Channels with large cross sections produce many counts and therefore have small uncertainties. Weaker channels accordingly are more uncertain.

The total cross section for charge transfer remains relatively flat over the presently considered energy range. Also, for example, we see from the tables for Ar¹⁸⁺ impact that capture to n=9 is the dominant channel, especially below about 60 keV/u. Above this, neighboring levels are also quite strongly populated. This is in good accord with previous results of Olson (PRA 24, 1726 (1981)) and Olson and Schultz (Physica Scripta T28, 71 (1989)) which showed that the n-level with highest probability should scale as q^3/4 where q is the projectile ionic charge (here q=18 and n_max = 8.7). We also see that capture to n-levels lower than n_max are more flat with energy than those with n> n_max.

Regarding the l distributions, for l less than n_max, the peak level is about slightly below the given n-level (e.g. for Ar¹⁸⁺ at 100 keV/u, the l distribution for n=5 peaks at l=3 and for n=7 it peaks at l=5), whereas for n-levels around n_max, it peaks more sharply just before l = n. For n > n_max, all l distributions tend to peak near the same large value (i.e. around l_max about q^3/4) and then drop off sharply (e.g. for Ar¹⁸⁺ at 60 keV, for n > 13, the distributions all peak at about n=12). For the lower collision energies, these cross sections for the non-dominant levels drop and these clear patterns for higher n-levels are harder to see.

State-selective charge transfer computed using the CTMC method for Ar^15-18+ + H: