The fact that most extrasolar planets found to date are orbiting metal‐rich stars lends credence to the core accretion mechanism of gas giant planet formation over its competitor, the disc instability mechanism. However, the core accretion mechanism is not refined to the point of explaining orbital parameters such as the unexpected semimajor axes and eccentricities. We propose a model that correlates the metallicity of the host star with the original semimajor axis of its most massive planet, prior to migration, assuming that the core accretion scenario governs giant gas planet formation. The model predicts that the optimum regions for planetary formation shift inwards as stellar metallicity decreases, providing an explanation for the observed absence of long‐period planets in metal‐poor stars. We compare our predictions with the available data on extrasolar planets for stars with masses similar to the mass of the Sun. A fitting procedure produces an estimate of what we define as the zero‐age planetary orbit (ZAPO) curve as a function of the metallicity of the star. The model hints that the lack of planets circling metal‐poor stars may be partly caused by an enhanced destruction probability during the migration process, because the planets lie initially closer to their central star.

In 2005 we suggested a relation between the optimal locus of gas giant planet formation, prior to migration, and the metallicity of the host star, based on the core accretion model and radial profiles of dust surface density and gas temperature. At that time, less than two hundred extrasolar planets were known, limiting the scope of our analysis. Here we take into account the expanded statistics allowed by new discoveries, in order to check the validity of some premises. We compare predictions with the present available data and results for different stellar mass ranges. We find that the Zero Age Planetary Orbit (ZAPO) hypothesis continues to hold after a one order of magnitude increase in discovered planets. In particular, the prediction that metal poor stars harbor planets with an average radius distinctively lower than metal rich ones is still evident in the statistics, and cannot be explained away by chaotic planetary formation mechanisms involving migration and gravitational interaction between planets. The ZAPO hypothesis predicts that in metal poor stars the planets are formed nearer their host stars; as a consequence, they are more frequently engulfed by the stars during the migration process or stripped of their gaseous envelops. The depleted number of gas giant planets around metal poor stars would then be the result of the synergy between low formation probability, as predicted by the core accretion model, and high destruction probability, for the ones that are formed.

The chemistry along the mass loss of Hot Jupiters is generally considered to be simple, consisting mainly of atoms, prevented from forming more complex species by the intense radiation field from their host stars. In order to probe the region where the temperature is low (T < 2000 K), we developed a 1D chemical and photochemical reaction model of the atmospheric mass loss of HD 209458 b, involving 56 species, including carbon chain and oxygen bearing ones, interacting through 566 reactions. The simulation results indicate that simple molecules like OH + , H2O + and H3O + are formed inside the region, considering that residual H2 survives in the exosphere, a possibility indicated by recent observational work. The molecules are formed and destroyed within a radial distance of less than 10 7 km, but the estimated integrated column density of OH + , a potential tracer of H2, is high enough to allow detection, which, once achieved, would indicate a revision of chemical models of the upper atmosphere of Hot Jupiters. For low density Hot Jupiters receiving less intense XUV radiation from their host stars than HD 209458 b, molecular species could conceivably be formed with a higher total column density.