Highly ionized species, such as CIV, N V, and O VI, are commonly observed in diffuse gas in various places in the universe, such as in our Galaxy's disk and halo, high velocity clouds (HVCs), external galaxies, and the intergalactic medium. These ions are often used to trace hot gas whose temperature is a few times 105 K. One possible mechanism for producing high ions is turbulent mixing of cool gas (such as that in a high or intermediate velocity cloud) with hotter (a few times 106 K) gas in locations where these gases slide past each other. By using hydrodynamic simulations with radiative cooling and non-equilibrium ionization (NEI) calculations, we investigate the physical properties of turbulent mixing layers and the production of high ions (C IV, N V, and O VI). We find that most of the mixing occurs on the hot side of the hot/cool interface, where denser cool gas is entrained and mixed into the hotter, more diffuse gas. Our simulations reveal that the mixed region separates into a tepid zone containing radiatively cooled, C IV-rich gas and a hotter zone which is rich in C IV, N V, and Ovi. The hotter zone contains a mixture of low and intermediate ions contributed by the cool gas and intermediate and high-stage ions contributed by the hot gas. Mixing occurs faster than ionization or recombination, making the mixed gas a better source of C IV, N V, and Ovi in our NEI simulations than in our collisional ionization equilibrium (CIE) simulations. In addition, the gas radiatively cools faster than the ions recombine, which also allows large numbers of C IV, N V, and Ovi ions to linger in the NEI simulations. For these reasons, our NEI calculations predict more C IV, N V, and Ovi than our CIE calculations predict. We also simulate various initial configurations and find that more C IV is produced when the shear speed is smaller or the hot gas has a higher temperature. We find no significant differences between simulations having different perturbation amplitudes in the initial boundary between the hot and cool gas. We discuss the results of our simulations, compare them with observations of the Galactic halo and highly ionized HVCs, and compare them with other models, including other turbulent mixing calculations. The ratios of C IV to N V and N V to O VI are in reasonable agreement with the averages calculated from observations of the halo. There is a great deal of variation from sightline to sightline and with time in our simulations. Such spatial and temporal variation may explain some of the variation seen among observations.
