The existence of trapped gases in the EETA 79001 shergottite meteorite that were apparently shock-implanted from the Martian atmosphere raises important questions as to the mechanism of gas implantation and whether the implanted gas has been mass fractionated. To study the phenomenon of shock-implantation of gases, we artificially shocked whole-rock and powder samples of a terrestrial basalt to pressures of 2--40 GPa in the presence of controlled gas mixtures ranging from 10-4 to 3 atmospheres. Argon, Kr, and Xe, and to a lesser extent Ne, were readily implanted into the silicate under a wide range of experimental conditions. As exemplified by Ar, the amount of implanted gas is linearly proportional to its partial pressure over a partial pressure range of 0.0001 to 0.1 atmosphere, the former value being similar to the partial pressure of Ar on Mars. The implanted gas showed no evidence of isotopic fractionation, and for Kr and Xe fractionation limits of 0.1% per mass unit could be set. The elemental composition of shock-implanted gas also closely resembles the ambient gas phase, except that Ne is only lightly retained and readily leaks from the samples after the shock. Stepwise temperature releases of gas implanted at 2, 5, 20, and 35 GPa indicate two or more lattice sites for the implanted gas that are possibly related to microcracks and other lattice defects. Activation energies, Q, for diffusion of shock-implanted gas from the samples increase dramatically with increasing shock; values Q of for Ar diffusion for the four shock levels 2, 5, 20, and 35 GPa are 7, 9, 14, and ~25 kcal/mol, compared to 30--45 kcal/mol for typical radiogenic Ar. Diffusion of Kr and Xe is considerably slower than Ar. The amounts of gas that would have been implanted with 100% efficiency were calculated from the measured porosities of the powder samples and were compared to observed abundances. The implantation efficiencies were approximately 0.5% at 2 GPa, 7% at 5 GPa, and greater than 50% at both 20 and 35 GPa.

These high gas implantation efficiencies occur for shock pressure where modest amounts of sample melting at grain boundaries begins, and they suggest that higher shock pressures generating substantial melting could not achieve appreciably greater efficiencies. These experimental data are consistent with shock-implantation of Martian gases without mass fractionation into shock-melted phases of EETA 79001; the progenitor of these melts may have been porous, fragmental material.