A global potential energy surface (PES) for the 1 4A′ spin-aligned state of Li3 is presented. The surface is constructed as a many body expansion of the potential which is the sum of pairwise additive two-body potentials plus a three-body term. The two-body potential is that for the a 3Σu+ state of the lithium dimer. It combines the most recent Rydberg–Klein–Rees potential available [A. Ross (private communication)] with well-known short and long range expansions and accurately reproduces all known experimental data. To obtain the three-body contributions, an ab initio PES was computed at 1122 points using full configuration interaction for the three valence electrons with an augmented Gaussian basis and the effective core potentials of Stevens, Basch, and Krauss [W. J. Stevens et al., J. Chem. Phys. 81, 6026 (1984)] for the other electrons. The two-body interactions are also calculated using the same basis and then subtracted from the full interaction to give the three-body term. To construct the three-body potential at arbitrary configurations we use interpolation for small perimeters of the triangle formed by the triatomic system and an analytic fitting function for large perimeters. A switching function guarantees the smoothness of the potential function everywhere. The equilibrium position occurs at D3h symmetry with a bond distance of 5.861a0, nearly 2a0 smaller than the equilibrium value of 7.886a0 of the lithium dimer. The well depth at the equilibrium is 4112.64 cm−1. This is considerably deeper than the well depth of 1001.22 cm−1 for the pairwise additive potential at its equilibrium. Three-body effects are even more important for Li3 than in the recently reported Na3 case [J. Higgins et al., J. Chem. Phys. 112, 5751 (2000)], and the nonadditive three-body term cannot be neglected in any calculation on this system.

Differential and angular integrated cross sections for elastic collisions of Ar atoms with Ar atoms are calculated from representative potential energy curves for collision energies from 0.01 eV to 10 keV. The scattering phase shifts are calculated using the WKB approximation. The cross sections are compared with data obtained from beam scattering experiments and transport coefficient measurements. On the basis of assessments of experimental reliability and of theoretical potentials, we make a selection from the available short-range interatomic potentials. Consideration of experimental elastic and inelastic cross section data leads to the suggestion that our theoretical results should be interpreted as the sum of the elastic and inelastic differential cross sections. The theoretical differential, total and viscosity cross sections are fitted with empirical functions covering the energy range from 1 eV to 10 keV to facilitate their use in models of discharge plasmas, etc. Procedures for use of these cross sections are summarized briefly.